ML18087A601

Safety Evaluation of PSEG Rod Exchange Methodology, Revision 1

ML18087A601
Person / Time
Site:	Salem
Issue date:	11/16/1982
From:	Blake R, Brown R, Rosenfeld E Public Service Enterprise Group
To:
Shared Package
ML18087A600	List: ML18087A599 ML18087A601
References
NFG-004, NFG-004-R01, NFG-4, NFG-4-R1, NUDOCS 8212010150
Download: ML18087A601 (34)

Text

..

Prepared by Reviewed by Approved by

(' *-

-~~~-~

• e SAFETY EVALUATION OF THE PSE&G ROD EXCHANGE METHODOLOGY R. A. Blake Nuclear Technology Engineer Nuclear DeparL~ent R. T. Brown Sen_:Or Engineer Nuclear Department Manager -

Nuc ear Fuel Nuclear Department a21201o150 -- a211*24*

, PPDR ADOCK 05000272 j'...

PDR NFG 004 Revision 1 Nov. 1, 1982

~#~

. /I-IS-82.

Date 11-1&. -BZ Date I /-/Cc-f"' 2-Date

I.

SECTION 1.0 2.0 3.0 4.0 s.o TABLE OF CONTENTS TITLE INTRODUCTION.

SUMMARY

AND CONCLUSIONS **

METHODOLOGY **** * **..

3 *. 1 3.2 3.3 Test Procedure.

Analytical Methods.......

Interpretation Procedure.

3. 3.1 3.3.2 3.3.3 Exchange Mode Rod Worths Dilution Mode Rod worths Design Verification ***

3.4 Measurement.Acceptance Criteria NFG 004 Revision 1 Nov. 1, 1982 PAGE

• 1

• . 2

• 4

• 4

• *

• 6

• *

• 8

• 9

• ~

9

.10

.16

3. 4.1

3. 4. 2 3.4.3 Dilution Measurement Criteria. *.16 Exchange Measurement Criteria.

.16 Remedial Action. * * * *

.18 EXPERIMENTAL VERIFICATION....

.21 REFERENCES *.....

...30

I_,

TABLE NO.

3.1 3.2

• ~ 3.3 3.4 4.1 4.2 4.3 4.4 4.5

-e NFG 004 Revision 1 Nov. 1, 1982 LIST OF TABLES TITLE PAGE ROD EXCHANGE NOTATION CONVENTION ******* 13 EXAMPLE OF RSE VERIFICATION PROCEDURE *...15 TRADITIONAL DILUTION MEASUREMENT ***

• .19 PSE&G ROD EXCHANGE TEST REVIEW CRITERIA...20 COMPARISON OF EXCHANGE MEASUREMENTS AND PSE&G CALCULATIONS *...*..******* 24 COMPARISON OF RSE DESIGN AND PSE&G CALCULATIONS.

• 2 5 DEVIATIONS BETWEEN EXCHANGE MEASUREMENTS AND RSE DESIGN CALCULATIONS ****..*.* 26 DEVIATIONS BETWEEN DILUTION MEASUREMENTS AND RSL~* DESIGN CALCULATIONS *. *. * * * *

• 27 EXPERIMENTAL VERIFICATION OF THE PSE&G ROD EXCHANGE TEST...*.***.*****

28 I

I*'

LIST OF FIGURES FIGURE NO.

TITLE 3.1 ROD EXCHANGE PROCEDURE.

4.1 CORRELATION OF FLUX REDISTRIBUTION NFG 004 Revision 1 Nov. 1, 1982 PAGE

.14 WITH DIFFERENCES IN.. RSE ERRORS **..**** 29 I

f.,)

NFG -

004 Revision 1 Nov. 1, 1982

1.0 INTRODUCTION

This.report describes a revised version of the PSE&G Rod Exchange Methodology.

The.original version of this methodology was successfully implemented for the Salem 1, Cycle 4 startup test, conducted in April of this year.

The revised procedures described in this report result from a review of the Cycle 4 test re-sults and discussions with the NRC.

The revisions include changes to the interpretation procedures described in Section 3.3.1.

These changes were made to incorporate comments received from the NRC.

Addi-tional changes were required to the Measurement Acceptance Criteria (Section 3. 4) to be 'consistent with the interpretation changes.

And, finally, there are changes to the Experimental Verification, Section 4.0, which reflect the effects of the revised inter-pretation procedure.

Section 3.1 of this report describes the mechanics of ihe plant test procedures.

" S1ection 3.2 describes ttie PSE&G core physics models and general calculational procedures used to generate the analyt:.:;al data used to infer the rod worths from the measurements described in Section 3.1.

NFU2/l Section 3.3 describes the procedures for interpreta-tion of the rod worths using.11easurement_s from Section 3*.1 and the analytical data from Section 3.2~ Key Notation conventions are defined in Table 3.1.

Section 3.4 describes the test review and acceptance criteria as well as the procedura for remedial action.

Section 4 presents the benchmark resul~s, which in-clude comparisons of dilution measurements, exchange measurements, and design calculations for Salem 1, Cycles 1 and 3.

1 of 30

2.0

SUMMARY

AND CONCLUSIONS NFG -

004 Revision 1 Nov. 1, 1982 The PSE&G Rod Exchange Methodology has been developed as a replacement for the boron dilution technique.

The exchange test can be implemented as a 10CFRS0.59 test change if it does not represent an unreviewed question.

The safety of the methodology has been re-viewed and the results are documented in this report.

A summary of this review is presented in this sec-tion.

Based on this safety evaluation, it is con-cluded that the exchange test does not represent an unreviewed safety question.

The PSE&G Rod Exchange Test Procedure represents a measurement technique not described in the FSAR.

According to the provisions of 10CFRS0.59, the licensee may perform such a test without prior NRC approval if it does not represent an unreviewed safety question.

A test would represent an unreviewed safety question if:

1.

The probability of occurrence or the consequences of an accident or. malfunction of equipment impor.:...

tant to safety previously evaluated in the Safety Analysis Report may be increased.

2.

A possibility for an accident or malfunction of a different type than any evaluated previously in the Safety Analysis Report may be created.

3.

The margin of safety as defined in the basis for any technical specification is reduced.

The. purpose of the rod worth measu~ements is to pro-vide verification that the Reload Safety Evaluation (RSE) is conservative with respect to the core shut-down capability.

In this context, the question of safety associated with the implementation of the exchange procedure is related to the degree of design verification provided, and the margin of safety main-tained during the procedure* execution.

The~e are NFU2/l 2 of 30

NFG -

004 Revision 1 Nov. 1, 1982

• .related to safety criteria 1 and 3 ab9ve.

The test does not introduce the possibility for a new type of accident or malfunction and, t~erefore, Criterion 2 does not apply.

For the purpose of performing a safety evaluation of the rod exchange test, comparisons are made to *the dilution technique in terms of the degree of design verification provided and the margin of safety main-tained during the-test execution.

The dilution procedure verifies the design (RSE) p~edictions of four (4) of the eight (8) Salem rod banks.

In contrast to this, the exchange,procedure verifies design predictions for all eight rod banks.

Since the exchange measurements are obtained directly from the dilution measurement of the reference bank, the degree of design verification provided by the exchange test for each bank is the same as would be obtained from the dilution procedure.

However, since the* exchange test measures all eight rod banks, it provides a greater overall degree of verification.

The equivalence** of th.e exchange and dilution verifica-

~ion accuracies has been experimentally verified by two separate -benchmark tests.

The safety margin parameters of concern during the execution of the rod worth measurement are the shut-down margin and ~he flux peaking factor.s.

The margins to limits associated with both of these parameters are significantly reduced with the insertion of rod banks.

The greater the number of rod banks inserted, the greaier the margin reduction.

NFU2/l The dilution measurement procedure requires the simul-taneous insertions of a minimum of four rod banks.

The exchange measurement procedure measures all eight rod banks but never requires the simultaneous inser-tion of more than two rod banks.

Therefore, signifi-cantly more margin is maintained during the execution of the exchange procedure than the dilution procedure.

3 of 30

1._i 3.0 METHODOLOGY

• e NFG -

004 Revision 1 Nov. 1, 1982 The PSE&G Rod Exchange Methodolqgy con~ists of four components; 1) the plant test or measurement proce-dure, 2) the analytical methods, 3) the interpretation procedure, and 4) the measurement acceptance crite-ria.

Each of these components are described in the following subsections.

3.1 TEST PROCEDURE The PSE&G Rod Exchange Test Procedure (1) consists of two steps:

First, the most worthy of the eight rod banks is choseri as a reference bank and is diluted frbm the full out to the full in (or nearly full in) position with all other rod banks remaining in the full-out position.

The worth of the reference bank is measured during this dilution using an on-line reactivity com~

puter.and standard data reduction techniques.

~he second step is to. perform a critical exchange between the reference bank and the bank to be meas-ured.** This is accomplished by withdrawing the refer-ence bank at constant boron concentration and tempera-ture and inserting the bank to be measured, referred to as bank x, in a manner such as to maintain the

• reactor nearly critical.

When bank x is fully in-serted, the position of the reference bank is adjusted to make the reactor just critical.

This just critical position is noted, *and the reference bank is then ex-changed with bank x in the opposite direction until the reference bank is again inserted and bank x withdrawn.

NFU2/l Another bank is then chosen for measurement, and the whole process of critical exchange is repeated.

Each bank is in this fashion "measured" against the cali-brated reference bank.

The measurement data consists 6f the absolute worth of the reference bank; and the 4 of 30

'*)

NFU2/l

-e NFG -

004 Revision 1 Nov. 1, 1982 relative worth of the other banks in terms of the critical position of the reference bank when displaced by the.measured bank.

These relative worths are con-verted to absolute bank worths using the Interpreta-tion Procedure described below in Section 3.3.1.

The above description of the rod exchange test proce-dure assumes that.initial and final reference bank positions represent exactly critical conditions and that there are no changes in moderator temperature or boron concentration during the test.

This, however, is an ideal situation and practical experience has demonstrated that small deviations from these ideal conditions are to be expected.

To account for these deviations, the test procedure includes steps to record the moderator temperature and core reactivity at each exchange end-point.

Also recorded are the measured differential rodworths for the reference bank in the presence of each of the bariks being measured (bank x = 0 steps).

Corrections are made to the.measured critical exchange positions of the reference bank based on the observed reactivity deviations and the measured differential rodworth

• 5 of 30

~

I I

NFG -

004

• Revision 1 Nov. 1, 198Z 3.2 ANALYTICAL METHODS The PSE&G analytical methods f o~ Rod Exchange Measure~

ments consists of a core model and a set of procedures for the application of that model.

PSE&G utilizes the ARMP(2) Code Package for the core model in all Rod Exchanges applications.

Since ARMP has become an industry standard code, no further description of the code package will be given here.

The PSE&G ARMP model of the Salem reactors represents a full core, three dimensional geometry with 12 axial nodes and one radial node per assembly.

This model is applied to a Rod Exchange Measurement for a given cycle by simulating both the Rod Exchange Test and the Standard Boron Dilution Test sequences.

The Standard Boron Dilution Test sequence is simu-lated by calculating the worth of each rod bank ifi' the sequential, nonoverlap insertion mode.

In this calcul~tion, Bank D is inserted first, Bank C is inserted next with D remaining in, Bank B is then inserted with D and c'remaining in, etc.

The RCS boron*concentration is varied during this simulation to maintain the core model nearly critical.

These bank worths are referred to as the "calculated dilu-tion mode" worths.

NFU2/l The Rod Exchange Test is simulated in three steps:

The first step in the simulation is to compute the worth of each bank with all *other rods out.

These bank worths are used to identify the reference bank.

second, the core reactivity is calculated as a func-tion of the reference bank position when the bank being measured, bank x, is fully inserted.

These calculations are used to predict the critical exchange reference positions.

6 of 30

• e NFG -

004 Revision 1 Nov. 1, 1982 3.2

• ANALYTICAL METHODS (Contd.)

NFU2/l The third step in the simulation is to compute the worth of each bank x, with the r~ference bank at the critical exchange position, all other rods withdrawn.

These are referred to as the "calculated or predicted exchange mode rodworths."

Since these calculations are performed prior to the test, the actual critical exchange positions are not known.

Therefore, several sets of calculations are performed over a range of reference bank positions to allow interpolation of the predicted exchange mode rodworths appropriate to the measured critical position.

7 of 30

- 1 1

. I

3.3 INTERPRETATION PROCEDURE

• e NFG 004

.Revision 1 Nov. 1, 1982 This section describes the procedure (3) for inferring the "measured exchange mode" bank worths from test data described in Section 3.1.

For the purpose of clarity, a set of rod Exchange Notation Conventions is introduced and used in the derivation of the below interpretation techniques as well as in later sectio.ns.

A typical rod exchange te~t maneuver begins with the core just critical, the reference bank nearly in-serted, and all other rods out.

The maneuver ends with the core again just critical, boron and tempera-ture unchanged, the reference bank partially with-drawn,. bank x (the* bank being measured) fully in-serted, and all other rods out.

This is illustrated in Figure 3.1 by the stair case path from point A to point E.

Since points A and E represent critical core conditions, the net change in core reactivity between them is zero.

This is not only true for path AE, but for any path beginning with A.and ending with E, such as ABE, ABCDE, or AFE.

Along any specific p~th, there

~re associated changes in core reactivity due to the motiops of the reference rod bank and bank x.

The reactivity change due to the motion of a rod bank is dependent on the particular path.

As an example, in Figure 3.1 paths AF, BE, and CD, all represent the motion of bank x moving from fully withdrawn to fully inserted.

The core reactivity associated with this motion differs for. each path due to the variable pres-ence of the reference bank.

For the purpose of dis-cussion, it is therefore desireable to have a notation convention for rod worth parameter¥ which i.dentif ies the path.

The rodworth notation convention used in this report is presented in Table 3.1.

This conven-tion identifies the path (l,m) and also ~he source of the data (s = calculated, measured via dilution, etc.).

T6is convention is used extensively in the following derivations and procedure descriptions.

NFU2/l 8 of 30

• 3.3.l EXCHANGE MODE ROD WORTHS

• e NFG -

004 Revision 1 Nov. 1,. 19 8 2 The actual rod exchang~ t~st procedure was described above in terms of Figure 3.1 as the path A E in which positions A and E represent critical conditions.

However, for the purpose of interpretation, the exchange test procedure can be imagined to be represented by the path A B E.

The reactivity balance equation for path A B E ca~ be writt~n as follows:

wexc

+

wexc AB,ref BE,x

=

0 (Equa 3-1)

The reactivity component due to bank x motion from B to E, wexc is defined as the BE, Exchange Mode Rod Worth for.bank x.

The component due to the reference bank motion, wexc can be determined directly from AB,ref the dilution measurement of the reference bank; wexc**

= - wdil

( Equa. 3-2)

AB,ref BA,ref substituting Equation 3.2 into 3.1, the exchange mode worth for bank x can be determined directly from the dilution measurement.

wexc BE,x

=

wdil BA,ref

( Equa. 3-2) 3.3.2 DILUTION MODE ROD WORTHS NFU2/l The dilution mode rod worth for any bank is that worth obtained via the path of seql;l_ential, non-overlap insertion from the all rods out condi-tion.

This path cannot be represented with a simple two dimensional graph such as Figure 3.1.

Therefore, the path will be referred to 9 of 30

NFG -

004 Revision 1 Nov. 1, 1982 3.3.2 DILUTION MODE ROD WORTHS (Contd.)

3.3.3 NFU2/l simply as the "dilution" path, using the subscript "dil" as explain~d in Table 3.1.

The dilution mode rod worth for any bank is generally greater than the exchange mode worth due to the presence of the other rod banks.

The actual dilution mode worth for any bank can be estimated from the measured exchange mode worth and analytical data as follows:

wexc

= wexc A cal dilx BE,x x

( Equa 3-4) where ;8 :al

= weal I

weal DIL,x BE,x DESIGN VERIFICATION PROCEDURE Rod worths measured via the dilution technique

• ~r~ directly related to the design values asso-ciated with the.Reload Safety Analysis (RSE). '

The traditional dilution acceptance criteria,

-*therefore provtde direct verification of design calculations.

The exchange mode rod worths measured using the exchange test are closely related to design values.

However, they require an intermediate evaluation step to properly verify design calculations.

It is the role of the PSE&G models to provide this intermediate step.

Design verification using th~ PSE&G models and the rod exchange test is accomplished in two steps.

First the measured exchang~ mode rod worths are compared to PSE&G exchange mode pre-dictions.

Second, PSE&G dilution mode predic-tions are compared to RSE design calculations.

consistency between ~he two sets of comparisons then constitutes verification of the d~sign 10 of 30

'.I 3.3.3 NFU2/l

• e NFG -

004 Revision 1 Nov. 1, 1982 DESIGN VERIFICATION PROCEDURE (Contd.)

calculations associated with the reload safety evaluation.

This procedure is detailed with an example in Tabie 3.2.

The data for this example was obtained from actual Salem 1, Cycle 3. test results.

The example addresses total rod worth, but the procedure can be, and is applied to all individual control rod bank worths as well as the total shutdown bank worth.

The exchange mode comparison described by Equation 3-5 in Table 3.2 relates the measured exchange mode worth to PSE&G model predictions.

This ia a direct comparison of measurement to prediction for all rod banks.

The Cycle 3 re-sults demonstrate that the PSE&G model predic-tions are 1.1% lower than direct measurements.

The dilution mode comparison described by Equation 3-6 relates the total rod worth pre-dicted by the RSE models to that predicted by PSE&G mod~ls.

Both_ predicted quantities repre-sent the dilution mode rod worths which are direct~j related to the reload safety evalua-tion.

The results shown demonstate that the PSE&G model predictions are 6.1% lower than RSE mode 1.

The RSE verification is described by Equation 3-7.*in Table 3.2.

As described in Section 3.3.1, the exchange measurements are essentially the same in nature as dilution measurements.

Therefore, it is approp~iateito obtain the RSE verification ratio as the product of the ex-change and dilution ratios as def i~ed in Equation 3-7.

The results shown in the example demonstrate that the RSE model predictions are too low, by 5.),% relative to measurements.

This result was confirmed by independent dilution 11 of 30

NFG -

004 Revision 1 Nov. 1, 1982 3.3.3 DESIGN VERIFICATION PROCEDURE (Contd.)

NFU2/l measurements which were m~de in parallel with the exchange test for the purpose of experimen-tal verification.

The verification results are presented in Section 4.

12 of 30

ws lm, x source TABLE 3.1 NFG -

004 Revision 1 Nov. 1, 1982 ROD EXCHANGE NOTATION CQNVENTION

=

examples, Change in core reactivity due to the movement of bank x along the path lm s

s =

s =

s =

s =

s =

act; cal; dil; exc; RSE; actual worth, not observable calculated worth, using models measured worth obtained via rod dilution test.

measured worth obtained via rod exchange test.

calculated worth using same model as those used for Reload ~afety Evaluation.

path examples, lm lm =dil; lrn = AE; bank examples, x

rod bank insertion con-f igu ration as required for sequential, nonoverlap insertion (dilution mode rod worth).

refer tolFigure 3.1, actual rod exchange test path A to E

ref =

reference bank x =

any bank, including the reference bank.

NFU2/l 13 of 30

I l

200 150 100 50 0

NFU2/l FIGURE 3.1

• e NFG -

004

.Revision 1 Nov. 1, 1982 R 0 D E X C H A N G E P R o* C E D U R E 0

I I ©-

so I.

100 REFERENCE BANK POSITION 14 of 30

--© I

200

• ~*....

.. )

Data TABLE 3.2 NFG -

004 Revision 1 Nov. 1, 1982 EXAMPLE OF RSE VERIFICATION PROCEDURE USING THE EXCHANGE TEST (Salem 1, Cycle 3 results)

(a)

Total dilution mode rod worth predicted from RSE models 3550 pcm

( b)

Total dilution mode rod worth predicted from PSE&G models 3332 pcm

( c)

Total exchange mode rod worth predicted from PSE&G models 2669 pcm

( d)

Total exchange mode rod worth measured from exchange test.

269 9 pcm Exchange Mode Comparison Exchange Ratio = Measurement (d)

=

2699

= 1.011 2669

( Equa 3-5)

PSE&G Cale ( c)

~his demonstrates that the measurements are 1.1%

greater than the PSE&G model predictions.

Dilution Mode Comparison*

Dilution Ratio = PSE&G Cale (b)

RSE Cale (a)

=

3332

=

0.939 3550

( Equa 3-6)

RSE verification

(

RSE )=

(Exch.)

ver~f Ratio Ratio NFU2/l This demonstrates that the PSE&G models predictions are 6.1% lower than RSE model calculations

• ( Di~ ) =.

\\Ratio 1.011

• 0.939 = 0.949 (Equa3-7)

This demonstrates that the RSE models are 5.1%

lower than measurements.

15 of 30*

J

• e NFG -

004

• Revision 1 Nov. 1, 1982 3.4 MEASUREMENT ACCEPTANCE CRITERIA The purpose of rod worth measurements following a reload is to provide a partial audit of the accuracy of the design calculations.

Acceptance criteria are therefore necessary to assure that the deviations between measurements and predictions are conservative with respect to the allowances for calculational uncertainty used in the Reload Safety Evaluation.

3.4.l DILUTION MEASUREMENT CRITERIA For rod worth measurements obtained by the dilution/boration technique, typical

~easurements include only the four control banks.

Typical acceptance criteria for deviations between measurements and design predictions are shown in Table 3.3.

3.4.2 EXCHANGE MEASUREMENT CRITERIA NFU2/l As described in. Secion 3.3.3, the design verification procedure associated with the rod

• • exchange test directly quantifies the differences between RSE design calculations and measurements~ Therefore, it is appropriate to apply the traditional dilution measurement acceptance criteria described in Table 3.3 to these verification results.

However, since the PSE&G model is used in the exchange test to provide an intermediate evaluation step, additional review criteria for model verification are required.

The PSE&G model is used in the design verification procedure only in a relative manner.

Observed differences between the model predictions and the exchange measurements are directly accounted for when verifying RSE rod worths.

For this reason, the accuraci of the verification procedure is largely insensitive to 16 of 30

3.4.2

~-

NFU2/l NFG -

004 Revision 1 Nov. 1, 1982 EXCHANGE MEASUREMENT CRITERIA (Contd.)

biases in PSE&G predictions.

However, it is reasonable to assume that the verification accuracy could be degraded given very large PSE&G model biases.

Therefore, additional review criteria for PSE&G model verification have been established for the rod exchange test.

These criteria are presented in Table 3.4.

The basis for these review criteria are presented below.

A review criteria of +/- 15% on individual banks is appropriate based on Salem 1, Cycle 1 test experience.

Exchange measurements of two shut-down banks (SD and SC) deviated by 15 to 16%

from PSE&G predictions as described in Section 4, Table 4.1.

However,* the.design verification accuracy for both banks were conf irrned by inde-pendent dilution measurements as described in Section 4.0.

A review ~riteria ot + 10% on the total bank worth is appropriate based on Salem 1 Cycle 1 test e:~erience and PSE&G model sensitivity studies.

Sensitivity studies were performed in which the PSE&G model was adjusted such that rod *worth for each rod bank was increased by 10%.

Predictions from the adjusted model were then used for Salem 1, Cycle 1 design verification evaluations.

These results were compared to evaluations made.

using the unadjusted model. 'The comparison demonstrated that the effect on design verif ica-tion accuracy was less than 1%.

The total predicted exchange mode worth for Salem 1, Cycle 1, differed from measurements by 10%.

(See Table 4.1).

However, the design verification accuracy for total bank worth was confirmed to better than 1% by independent dilu-tion measurements.

17 of 30

NFG -

004 Revision 1 Nov. 1, 1982 3.4.3 REMEDIAL ACTION NFU2/l Should the test results f~il to meet the accept-ance criteria of Table 3.3, the reactivity worth of control banks D through A shall be measured by the dilution/boration technique.

The accept-ance criteria presented in Table 3.3 shall be applied to the dilution measurements.

Should the results of the rod exchange test fail to meet the review criteria of Table 3.4, the Station Operations Review Committee will be informed.

The t~st results shall be reviewed by the committee and final resolution shall be based on the composite of plant start-up data

• and an evaluation of* the impact of the discrep-ancy on the results of the analyses of the applicable events considered in the FSAR.

18 of 30

/.

NFU2/l

1)

2)

TABLE 3.3

• e NFG -

004 Revision 1 Nov. 1, 1982 TRADITIONAL DILUTION. TEST ACCEPTANCE CRITERIA Individual control and shutdown bank worth IL%1 15% for banks > 600 pcm 16%1

< 100 pcm for banks < 600 pcm Total of all Control and Shutdown Banks 16%1 10%

Definition

6. % = measured -design design

~

x 100%

19 of 30

NFU2/l TABLE 3.4 NFG -

004

• Revision 1 Nov. 1, 1982 PSE&G ROD EXCHANGE ~EST REVIEW CRITERIA

1)

Individual control and *shutdown bank worth 16%1

< 15%

for banks > 600 pcm 16%1

< 100 pcm for banks < 600 pcm 2).

Total of all Control and shutdown Banks Definition

= measured. -de~ign design 20 of 30 x "10.0%

!i

NFG -

004 Revision 1 Nov. 1, 1982 4.0 EXPERIMENTAL VERIFICATION The design verification procedur;e using the rod exchange procedure was described in Section 3.

This procedure quantifies the error in dilution mode rod worths predicted by the RSE design models using. the exchange measurements and the PSE&G models.

This procedure has been verified by independent dilution measurements.

This verification is described in this section.

Rod exchange measurements were made during the startup tests of Salem 1, Cycles 1, 3, and 4.

Parallel dilu-tion measurements* were made for Cycles 1 and 3, but not for Cycle 4.

The results of the exchange measure-ments and the corresponding PSE&G exchange predictions are presented in Table 4.1.

Also.shown are the ex-change mode ratios which have* been calculated in accordance with the design verification procedure described in Table 3.2, Eauation 3-5.

The rod worths predicted by the RSE models and compar-isons to PSE&G predictions are presented in Table 4.2.

+.hese d_ata correspond to the rod banks measured

• via the exch~hge test which were presented in Table 4.1.

Also shown in Table 4.2 are the dilution mode ratios which have been calculated in accordance with the *design verification procedure described in Table 3.2, Equation 3-6.

The exchange and dilution ratios calculated in Tables 4.1 and 4.2 respectively are multiplied in Table 4.3 to obtain the RSE verification ratio in accordance with Equation 3-7 of Table 3.2.

The results provide a ditect measure of the errors in the predicted RSE rod worths.

NFU2/l 21 of 30

I J

J I

4.0 EXPERIMENTAL VERIFICATION (Contd.)

• e NFG -

004

• Revision 1 Nov. 1, 1982 The RSE errors obtained in Table :4.3 were obtained

• .using the exchange measurements and the PSE&G models.

These results can be verified by.independent dilution measurements.

The difference between the dilution measurements and the RSE predictions should match the results in Table 4.3.

Differences between the RSE errors obtained by dilution and those obtained by exchange must be attributed to the combined uncer-tainties of the dilution and exchange measurement uncertainties.

The comparisons of the dilution measurements to RSE predictions are presented in Table 4.4.

The apparent RSE errors are calculated in a manner consistent with the apparent errors obtained from the exchange measure-ments presented *in Table 4.3 The two sets of errors are compared in Table 4.5.

The difference in the.

apparent errors are approximately +5% for Cycle 1 and'+

0.5% for Cycle 3.

No dilution measurements were made -

for Cycle 4.

The larger apparent differences' for the, Cycle 1 measurements are attributable to dilution measurement errors.

The cause of these errors was dis-cover~d after the cycle start up tests were completed, and were eliminated prior to performing the cycle 3 start up tests.(5).

The Cycle 1 dilution measurement errors were determined to be due to the effects of spatial flux redistribution caused by the rod motions during the dilution test.

During the dilution measurement of a given rod bank, the insertion of that bank may cause a radial redistri-bution of the neutron flux.

This r~distribution may result in a net increase or decrease in the relative flux in the fuel assemblies near the core:periphery.

This in turn would cause a change in the ex-core detec-tor signal which feeds the reactivity computer.(4).

Since the computer cannot distinguish between signal changes due to reactivity NFU2/l 22 of 30

r( l J

.J l

4.0 EXPERIMENTAL VERIFICATION (Contd.)

-e NFG -

004 Revision 1 Nov. 1, 1982 effects and those due to spatial effects, an error would be induced in the indicate'd rod worth.

Redis-tribution which causes a net increase in ex-core sig-nal due to rod insertion is referred to as positive redistribution.

Positive redistribution will cause the indicated rodworth to appear too small.

In con-trast, negative redistribution will cause the indi-cated rod worth t6 be too large by decreasing the ex-core signal.

The redistribution effects due to rod bank insertions during the Cycle 1 test have been estimated from model calculations and the results are presented in Table 4.5.

The values represent the per-cent change in ex-core sig~al due only to redistribu-tion during rod bank motion.

The sign of the changes are consistent with above definitions; ie, positive changes cause positive errores.

It is apparent that there is a very strong correlation between the redis-tribution and the differences in the apparent RSE err"ors.

A graph of this correlation is presented in Figure 4.1.

The spatial redistributions were still present during

.. Cycle 3 meast~ements, but their effect on the reactiv-ity computer was eliminated by changes in the test procedure. (5).

NFU2/l 23 of 30

t".

Cycle Cycle 3 Cycle 4 NFU2/l TABLE 4.1 NFG -

004

.Revision 1 Nov. 1, 1982 COMPARISON OF EXCHANGE MEASUREMENTS AND PSE&G CALCULATI.ONS Measured PSE&G EXCHANGE BANK wexc weal RATIO x

exc x exc x (Meas)

(pcm)

(pcm)

(PSE&G) 1 D

1107 1031 1.074 c

825 741 1.113 B

522 467 1.118 A

924 858 1.077 SD 469 403 1.164 SC 351 305 1.151 TOTAL 4198 3805 1.103 D

834 859

. 9 71

• c 678 640 1.059 B

384 387

. 9 9 2 A

803 783

1. 026 TOTAL 2699 2699 1.011 D

862 847 1.018 c

598 595 1.005 B

370 410

• 90 2 A

8 20 799

1. 026 SD 266 3!5 SC 285 271

• 9 7 2 SB 610 625 SA 745 749 TOTAL 4556 4611

• 9 82 24 of 30

r<.

)

Cycle 1

Cycle 3 Cycle 4 NFU2/l TABLE 4.2

• e NFG -

004*

Revision 1 Nov. 1, 1982 COMPARISON OF RSE DESIGN AND PSE&G CALCULATlONS RSE PSE&G DILUTION BANK WRSE weal RATIO x

dil,x dil,x (PSE&G)

(pcm)

(PCID)

(RSE)

D 1076 1031

• 958 c

1014 1011

• 997 B

770 706

• 917 A

1155 1107

• 958 SD 7 25 659

.909 SC 1183 1081

• 914 TOTAL 5923 559 5

• 9 4 5 D

ass*

859

.971 r"'

953 908

• 9 5 3 B

633 566

.894 A

1079 999

• 9 26 TOTAL

sso 3332

. 93 9 D

929 847

• 912 c

8 08 800

.990 B

592 577

.97 5 A

1137 11123

• 9 8 8 s

3092 2834

• 91 7 TOTAL 6558 6181

• 943 25.of 30

r _,r I..

)

BANK x

Cycle 1 D

c B

A SD SC

• • TOTAL Cycle 3 D

c B

A TOTAL Cycle 4 D c B

A s

TOTAL NFU2/l NFG -

004

Revision 1 Nov. 1, 1982 TABLE 4.3 INFERRED DEVIATION BETWEEN DILUTION MEASUREMENTS AND RSE DESIGN CALCULATIONS RSE Exchange Dilution Verif.

Ratio Ratio Ratio meas PSE&G

=

meas PSE&G x

RSE RSE 1.074 0.958 1.029 1.113 0.997 1.110 1.118 0.917 1.025 1.077 0.958

1. 032 1.164 0.909 1.058 1.151 0.914

1. 052 1.103 0.945 1.042 0.971

0. 971 0.943 1.059 0.953

1. 009 0.992 0.894 0.887 1.026 0.926 0.950 1.011 0.939 0.949 1.018

0. 912 0.928 1.005 0.990

• 0.995

.902 0.975

. 0. 8 79 1.026 0.988 1.014

.972 0.917 0.891 0.982 0.943 0.925 26 of 30 APPARENT ROD WORTH RSE ERROR (meas-RSE)

+ 2.9

+11.0

+ 2.5

+ 3.2

+ 5 *-.8"

+ 5.2

+ 4.2

- 5.7

+ 0.9

-11.3

- 5.0 5.1

- 7.2

- 0.5

-12.1

+ 1.4

-10.9

-7.4

( *"

' " )

NFU2/l 27 of 30 NFG.,.. 004 Revisi.on 1 Nov. 1, 1982

Bank x

Cycle 1 D

c B

A SC SB TOTAL Cycle 3 D

c B

A TOTAL NFU2/l TABLE 4.5 NFG -

004

• .Revis ion 1 Nov. 1, 1982 EXPERIMENTAL VERIFICATION OF THE PSE&G ROD EXCHANGE TEST APPARENT RSE ERRORS DIFFERENCE Flux Redistribution Exchange Dilution

( Dil-Exc) during Measurem' t.

Measurern't.

Dilution Measurements

+ 2.9

+ 2.9 o.o

- 13

+11.0

+16.7

+ 5.7

.f. *. 6 0

+ 2.5 0.5 3.0 24

+ 3.2

+ 7.5

+ 4.2

+ 40

+ 5.8

+ 2.8

- 3.0 20

+ 5.8 0.2 5.4 36

+ 4.5

+ 5.* 1

- 0.4 5.7 5.7 o.o

+ 0.9

+ 0.7 0.2

-10.3

-10.7

+ 0.4

- s.o

- 5.2 0.2 5.1 4.7

+ 0.7 28 of 30

t*:

-~

.-50%

NFU2/l FIGURE 4.1 NFG -

004 Revision 1 Nov. i,* 1982 CORRELATION OF FLUX REbISTRIBUTION WITH DIFFERENCES IN RSE ERRORS 5%

o*

e-1--

/

I

/

I I

B/

80sc T

T I I I

-i i

l Observed

• Differences

/

c

-#----l---j-----1-----1-----.__1 -

50%

i ! I I -I I T T-5%

Flux Redistribution 29 of 30

• I

5.0 REFERENCES

NFG -

004

• Revision 1 Nov. 1, 1982 (1)

Rod Swap Reactivity Measure~ent Method Test, Reactor Engineering Manual, Part 10.

(2)

ARMP:

Advanced Recycle Methodology Program EPRI Research Project 118-1, September 1977.

(3)

Rod Exchange Inferencing Procedure, NFT Design and Licensing Procedure DOl.6-03004.

(4)

Westinghouse Solid-State Reactivity Computer Manual, Westinghouse Nuclear Energy Systems, July 1974.

(5)

NAI Rod Swap Test Procedure, NAI 80-73.

NFU2/l 30 of 30