ML20065H955

Forwards Response to Core Performance Branch 820816 Concerns Re Methods of Monitoring Core Power Distributions for Cycle 7 Operation

ML20065H955
Person / Time
Site:	Prairie Island
Issue date:	09/29/1982
From:	Musolf D NORTHERN STATES POWER CO.
To:	Office of Nuclear Reactor Regulation
References
NUDOCS 8210050319
Download: ML20065H955 (13)

Text

_ _ _ _ _ _ _ _. _ _ _ _ _ _ _

v Northem States Power Company 414 Neollet Man Minneapoks. Minnesota $5401 Telephone (612) 330 5500 w

September 29, 1982 Director Office of Nuclear Reactor Regulation U S Nuclear Regulatory Commission Washington, DC 20555 PRAIRIE ISLAND NUCLEAR GENERATING PLANT Docket Nos. 50-282 License Nos. DPR-42 50-306 DPR-60 Response to NRC Concerns on Monitoring of Core Power Distributions in Prairie Island 1, Cycle 7 gj i

On August 16,1982 (and subsequent dates) discussions were held between NSP and the Core Performance Branch on methods of monitoring core power i

distributions. In response to staff concerns, the attached response is A

submitted for information.

Please call us if you have further questions concerning this subject.

kb s David Musolf Manager of Nuclear Su rt Services DMM/TMP/js NRR Regional Admin-III NRR Proj Mgr, NRC NRC Resident Inspector G Charnoff D

8210050319 820929 PDR ADOCK 05000282

.P PDR

RESPONSE TO NRC CONCERNS ON MONITORING OF CORE POWER DISTRIBUTIONS IN PRAIRIE ISLAND 1, CYCLE 7 Ref:

1.

Letter from J S Holm (ENC) to R Anderson (NSP), August 13, 1982, "Use of the Symmetry Option in the Detector Routine of the Conform Package for Prairie Island 1, Cycle 7" 2.

NSP Topical Report NSPNAD8101P, " Qualification of Reactor Physics Methods for Application to PI Units", December 1981.

Reference 1 is Exxon Nuclear's (ENC) justification for using the symmetry option of the DETECTOR code for monitoring core power distributions for PI1, Cycle 7.

The main point of their letter is that the symmetry option of the DETECTOR code can be used for core monitoring as long as core symmetry can be shown to be within some small degree uncertainty. They show this by comparing the differences between symmetric measured reaction rate locations. The option is acceptable as long as symmetry can be shown.

It should be noted that all measured reaction rates are used in determining core power distributions using this option.

For more details on the justification, see Reference 1 attached.'

The NSP DPS program, described in detail in Reference 2, was also used to analyze the core power distributions in PII, Cycle 7.

The DP5 program calculates reaction rates, power distributions, temperature distributions, etc.

for all assemblies in the core (full core representation). The calculation is done in three dimensions, thus incorporating all feedback effects. The program

.is normalized at BOC to a two-dimensional core PDQ depleted over the cycle length. Comparisons have been made at various state points over the cycle between calculated DP5 reaction rates and monitored reaction rates (raw data 1.e. the DETECTOR code has not. manipulated or calculated this data). Figure 1 gives the results of this comparison for MAP 107-27, the mean 1s 1.24 and the standard deviation is 1.01.

Figure 2 gives the comparison of the-DP5 calculated ~

~

FAH t the DETECTOR calculated F using the symmetry option. The technical AH specification' limit is also indicated.

The uncertainty associated with the DP5

r w

calculation of F is well established (see Reference 2) and is given as

.06 3g of the calculated value. This uncertainty includes uncertainty in calculation as well as in measurement.

(When this uncertainty was established over several cycles of operation, the measurement uncertainty was never subtracted out).

Figure 3 shows a comparison of the DP5 calculated Fgg's to the DETECTOR calculated F

's using the ' nearest neighbor option.'

AH The 095 program accounts for asymmetries in assembly exposure distributions only. Other asymmetries such as flow induced, etc. are not accounted for.

Therefore a significant deviation between the comparison of the calculated to measured reaction rates in a few monitored assemblies would indicate an asymetric core.

The comparisons shown in Figure 1 are all within 1% to 2% in the interior of the core thus indicating that the core is symmetrically loaded which corroborates the ENC conclusion.

Furthermore, the DP5 program calculates the power distributions in the unmonitored assemblies to within the same degree of accuracy as it does for the monitored assemblies.

In this mode, DP5 is calculating the power in unmonitored assemblies in a much more accurate but similar method to the DETECTOR code ' nearest neighbor option.' The coupling factors are inherent in the basic methodology of DP5, whereas for DETECTOR, they are explicitly determined and applied in an algorithim to distribute power to the unmonitored assemblies with no feedback effects accounted for.

Plant procedures require that DETECTOR and DP5 are to be run to determine a peak F AH for the core. For conservatism, the largest value from either code woulo be used for compliance. Tne DP5 program will be run at the plant conditions at the time the flux map, which was input to DETECTOR, was run.

The results from DP) are compared to the measured reaction rates. Acceptance criteria is then applied to this comparison. As long as the DP5 data is within the acceptance criteria, the DP5 results can be used to determine the peak F in m nitored and unmonitored AH assemblies.

It should be noted that DP5 used in this manner is completely independent of the DETECT 0,R code.

6-l

7-i i

A simpler way to use DP5 in the monitoring mode would be to use the measured reaction rates as input to DP5 and from this calculate assembly power distributions, in a sense, run DPS backward.

Running the program in this manner would eliminate the need to make the comparison outlined above. This would be the only difference between the two methods of running DP5 in the monitoring mode.

In the near future the Nuclear Analysis Department intends to produce a topical report on this mode of running DP5 for NRC review. When this review.is complete and the plant has reviewed and approved these methods, then NSP intends to use DPS as the monitoring tool for technical specification compliance for both Prairie Island plants.

L

Measured and Calculated Integrated Detector Responses Pl 1 CYCLE 7 HFP,10.106 GWD/MTU, ARO, EO XENON Absolute Differences FIGURE 1

MAP 107-27 1

2 3

-4 5

6 7

8 9

10 11 12 13 A-Abs. Dif f = (Meas - Calc)

• 1oo

~

d'J U[K.h (o.eJ C-(o.eJ o

1 J o.

c.p rm (aJ (J

(3 3

(h

()

E-g (j

(y F-(o.4J (o.2>

9'J

(

G-Co.e C.e 9.'7 (s.o (1.2>

(J U

(a.7) 1 U

u H-(1.33 (1.73 (0.1)

(o.s3

~

q qy y

j c m'y r'3 cm 1-e-

cy c5m J-

< s>

<1.2>

<og>

qo.8j gy qy q

K-o.)

2 u

L-b.o

-8 Pin Gd O

V.7 1

-12 Pin ad Detator t.oation -

M-(4.)2 x.x with

....F H Abs. Dif ference

)

3 PRAIRIE ISLAND 1 CYCLE 7 l

FAH MEAS VS CALC l

FIGURE 2 2

1.9 - - + -.-- -+ -- + --

+-

+ - -

+ -

- i- -

-i-- :

=-

1g_.....

1 lT.S. LIMITl:!

i l

1.7 -

r--

--v-

r. "

1.6 _

l 4

g-A 15-m p

~

v w

n.

S.

. f....{: '\\"'::!*~:.s,.a:: :g:i:: :&......j.h:R.R:$. =gw.;...f..'O. ;,....!....Q :,j~~. ~ Q.i.,..}.

~

1.4 -

W.

• y I

1.3_...:......:..... :.

Legend l

i i

12-

.-~ ~ -.- ~-

"~"

-~-~~--""--*-~~.---~-~~'~:'.~-

UPPER BOUND

.F.A...H.. CALCULAT.E..D..

1.1 -...:...........:..........:..............:........... :.. :........ :... :..:.......

LOWER BOUND i

c SYMMETRY OPTION j

T. S. LIMIT se l

l l

l l

l I

I I

l l

l l

l l

l l

l l

l l

l l

l l

l 0 0.5 1 1.522.533.544.555.566.577.588.599.51010.51111.51212.51313.5 EXPOSURE (GWD MTU)

NSP-NUCLEAR ANALYSIS DEPT.

PRAIRIE ISLAND 1 CYCLE 7 FAH MEAS VS CALC FIGURE 3 2.l l

1.9 -................

1.8_

1.7_

1.6_

A 15

......n..

~.

. e.

- :.N si h.

,.,,,.4...t.a. h. ~h.............*t....b...... ; e....,.... y...... %.... b..

f i

h

....!.... g

7...

.i

},4

.=.*y.

...s s

...p

-W

. Y.

t h.

1. -..~.

... ~............

oil 1.2 _

UPPER BOUND

.F. AH.. C..A..LC.ULATED.

1.1_................

LOWER BOUND i FJII MEASURED T. S. LIMIT eq 1

i i

i a

i i

i a

iii i

i i

e i

i i

i i

i i

s 0 0.5 1 1.522.533.544.555.566.577.580.599.51010.51111.51212.51313.5 m

EXPOSURE (GWD MTU)

NSP-NUCLEAR ANALYSIS DEPT.

E$N NUCLEAR COMPANY,Inc.

1131/dreAqadrBasd P. O Bau 1.kl Addsed Washrspan.9GE2 N Phone:(509)375-8'00 Telex: 15-2878 August 13, 1982 File LB/JSH:020:82 Mr. Roger Anderson Gereral Superintendant of Nuclear Analyses

• Northern States Power 414 Nicollett Mall Plaza Minneapolis, MN 55401

Subject:

Use of the Symmetry Option in the DETECTOR Routine of the CONFORM Package for Prairie Island Unit 1 Cycle 7 Reference 1:

Letter to Matt Klee (NSP) from MR ".111 gore (ENC) July 22, 1982.

Reference 2:

J. S. Holm, " Exxon Nuclear Analysis of Power Distribution Measurement Uncertainty for Westinghouse PWRs", ENC, July 1979.

In Reference 1 it is recommended that the symmetry option in DETECTOR be utilized in the monitoring of the core power distribution in Prairie Island Unit 1 Cycle 7.

This recommendation has been made based upon the fact that ENC believes that the DETECTOR option currently utilizied in the monitoring of the Prairie Island Unit 1 Cycle 7 core power distribution is overly conservative. The use of the symmetry option in DETECTOR will result in a more accurate measurement of the peak Fay in the core for Cycle 7.

The supportin'g argument for this recommendation is provided below.

The suitability of utilizing the symmetry option depends on the assumption of care quadrant symmetry with regard to the core assembly power distribution.

The measurement of the assembly power in any one of four symetric core locations provides a measurement of all four locations when quadrant symmetry exists in the core. To demonstrate that the use of the symmetry option for Prairie Island Unit 1 Cycla ' 1s suitable a comparison of

August 13, 1982 Page 2 LB/J5H:020:82 one-eighth core measured reaction rates in symmetric locations has been made. One-eighth core symmetric comparisons are made since tiiere are few one-quarter core symmetric measured locations.

In Figure 1 the differences between the measured and calculated reaction rates in the unrodded plane are shown.

In Figure 2 these differences have been folded into an eighth core map to compare the differences between calculated and measured powers in symmetric locations. The one-eighth core values are subtracted and the results shown in Figure 2 to represent the difference 1

t,etween measured reaction rates in symmetric locations with the expected eighth core asymmetries removed. The maximum difference between symmetric j

measured reaction rates as shown in Figure 2 is 1.9% with a standard m

deviation of 0.93%. The relative standard deviation is calculated assuming Y"

that each differenc6 represents two data points, one positive and one negative.

In Reference 2, the relative standard deviation with regard

=

to the measurement of a single assembly reaction rate is shown to be

. 6 5 %'.

The expected relative standard deviation for a symmetric core withregardtothegigrencebetweentwomeasuredreactionratesis, therefore, (2 x.65 )

=.92%.

The relative standard deviation for the differences between symmetric measured reaction rates shown in Figure i

2 is consistent with the expected relative standard deviation for a symmetric core.

N The uncertainty in F (and similarly F as calculated by DETECTOR using the symmetry op$1on can be estima9e)d from the data.in Reference 2 i

3 for cores which have symmetric power distributions.

The peaking factor F

has three components when determined by DETECTOR using the symmetry 39 option.

The three components are the measured reaction rate, the calculated j

reaction rate to assembly power ratio and the calculated local peaking factor. F is defined (with suitable normalization) as the measured d

3 reaction ra$e divided by tne ratio of the calculated reaction rate to assembly power ratio and multiplied by the calculated local peaking factor. The uncertainty in Fag as determined by DETECTOR using the symmetry option can be expressed as a combination of the uncertainties 1

in the components of Fgg.

Using the nomenclature of Reference 2, the

-l uncertainty in F3g, S m, it defined as:

l pxy J

/

)

S m /P*

=I (S R[ + (S

• I y) + (S c/

/

APR

, A UI p

y

\\

xy xy L

I

.,f

• l l: "

- - - - - - - - - - - - - - - - - - - - - ' - - - - - - ' - - - - - - - - ~ ~ - ' -

. August 13, 1982

-Page 3 LB/JSH:020:82 where S m /P" Relative standard deviation in F

=

p y

AH xy S m /A Relative stande.rd deviation in the measured

=

A

• Y xy assembly reaction rate S

/APR APR Relative standard deviation in the ratio of the

=

calculated reaction rate to assembly power l

S c/L Relative standard deviation in the local power

=

L peaking factor The values determined in Reference 2 for S * /^ y and S c/L are.65%'

Axy L

and 1.35%, respectively. The value of S

/APR can be determined from APP Equation 4.7 in Reference 2 for S /CF (S /CF = 2.058).

CF CF h l/2

/

I /R)2 + 2 (S

/APR)2.

2)

S l I

CF R

APR where S /CF CF Relative standard deviation in the ratio of the

=

calculated reaction rates in two assemblies I

S /R R

Relative standard deviation in the ratio of the

=

calculated assembly power in two assemblies.

w kr r.

a-

r:

~

August 13, 1982 Page 4 LB/JSH:020:82 4

it is reasonable to assume that the uncertainty in the ratio of the calculated assembly powers between two assemblies, S,/R, is greater than the uncertainty in the ratio of the calculated reactkon rate to calculated assembly power in one assembly.

AconservativeestimateofS[R

/APR I"

9"# I "A can be obtained by substituting SAPR ^

R S

/APR = (S /CF)2/3)1/2 = 1.19%.

APR CF The relative standard deviation for FAH' P* I can be determined y

xy from Equation 1 gith the valges 1.19%,.65% and 1.35% substituted for S

/APR, S * /A APR A

q ""d 3 c/L, respectively.

The resultant value for xy L

S m /P" is 1.91%. The one sided 95-95 tolerance limit is then 1.72

• pxy 1.91% = 3.29% assuming appropriate for S m /P,that the tolerance limit factor in Reference 2 is defined by Equation 1.

pxy The current Technical Specification value for the uncertainty.in F is applicabletotheuseoftheDETECTORcodewiththesymmetryoptiokH since 3.29% is less than'4.0%.

Sincerely, N

J. S. Holm, Unit Manager BWR Neutronics JSH/ mar cc: Matt Klee OH Peterson LC O'Malley Cliff Bonneau V RB Stout GA Sofer FB Skogen

1 Figure 1 Prairie Island Unit 1 Cycle 7 Map 107-27 Relative Reaction Rate Differences Configuration 2 only - 10,100 MWO/MT M

L K

J I

H G

F E

D C

B A

4.4 I

1.4

-4,9 2

5.8

-0.1 4.2 3

4.2 6.2

-5.0 4

-2.0

-5.4

-3.5 5

-5.1 NA 3.0 6

-4.2

-0.5 3.1 7

-1.4

-2.5

-1.9

-0.9 2.5 8

-3.5 3.7 1.3 g

-2.3

-2.8 NA 10 0.7 0.3 4.4-11 5.0 3.7 12 4.2 13 7s 3, 2.

0~ 5 1, 7

G F

E D

C B

A

-5.1(3) 3.1(2)

-0.5(2) 1.4(1)

-4.2(2) 7 NA 3.7(3) 0.3(3)

NA

-4.2(3) 0.6 0.8 0.0

-1.9(4)

-5.4(1)

-1.4(3)

-0.9(4) 3.0(1) 2.5(4) 8

-2.5(3)

NA 02.8(3)

-0.1(2) 3.7(4) 4.4(2)

-6 1.4 0.8 0.7 1.9

-2.0(2)

-3.5(1) 1.3(4)

-3.5 9

NA' O.7(3)

-4.9(3)

(1) 0.6 1.4

-2.3(3) 6.2(2) 4.2(2)

Reaction Rate 10 5.8(2) 5.0(3)

Differences (Quadrant) 0.4 0.8 Absolute Difference Between Reaction 4.4(4)

Rate Differences d 11 4.2(1) 4

.2 12 N

2

[}dj/N]

,. =

1

.93

=

13 Quadrant Key 2

1 3

4 Figure 2 Prairie Island Onit 1 Cycle 7 Map 107-27, 10,100 MWO/MT Relative Reaction Rate Differences Folded into an Eighth Core l

.