ML20082H124

Unit 1,Cycle 3 Startup Test Rept, Dtd Jul 1991

ML20082H124
Person / Time
Site:	Braidwood
Issue date:	07/31/1991
From:	Checca A, Gurley J, Jesus Sanchez COMMONWEALTH EDISON CO.
To:	Davis A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
References
NUDOCS 9108230097
Download: ML20082H124 (43)

Text

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

s -. Ccmmonw::alth Ediscn 1400 Opus Place

, { b'~~ ' Downsrs Grove, Illinois 60515 August 19, 1991 Hr. A. Bert Davis Regional Administrator i U.S. Nuclear Regulatory Commission l Region III  :

799 Roosevelt Road Glen Ellyn, IL 60137

Subject:

Braldwood Station Unit 1 Cycle 3 Startup Report NRC_Docketlo.,_50-456

Dear Mr. Davis:

Section 6.9.1 of the Braldwood Technical. Specification-states, in addition to the applicable reporting requirements of Title 10, Code of Federal Regulations, the following reports shall be submitted to the Regional Administrator of the NRC-Regional Office unless otherwise noted:

6.9.1.1. A summary report-of plant startup and power escalation testing _shall be submitted following: (1) receipt of an'0peration License,-(2) amendment to the. license involving a planned increase.In power level, (3) installation of fuel that has a different design or has been manufactured _by a-different fual supplier,;and_(4)_

modifications-that may.have significantly altered the nuclear, thermal,--or hydraulic performance of the plant.

6.9.1.2. The Startup Report shall-address each of the tests-Identified in the Final Safety Analysis Report FSAR and shall Include a description of the. measured values of the: operating conditions or characteristics obtained during.the test program and a. comparison of these values with design _ predictions-and-specifications. _Any corrective actions-that-were required.to-obtain_ satisfactory-

. operation sha11Lalso be described. Any-additional specific details required in license condittons. based on other: commitments shall'be-included in this report.

i6.9.1.3. :Startup Reports shall be submitted within: !(1)190 days following completion =of the-Startup Test-Program,.(?) 90? days.

following resumption or commencement of: commercial powerLoperation,-

>or (3) 9 months following initial criticality, whichever Is-earliest. If.the Startup Report:does'not cover all:1three events-(i.e;, initial? criticality,'completionfof Startup Test Program, and resumption of: commencement of commercial ~ operation)_ supplementary-L reports shall be submitted at least every-3. months unt11 al1 three-.

events have-_been. completed; h

w oc[ oN$ 6 T6M op . .

,' 9 p Hr. A. Bert Davis August--19,~1991-In accordance with Specification.'6.9.1.1(3) and 6.9.1.3(2), enclosed is the Braidwood Unit 1, Cycle 3 Startup: Report. Also enclosed, in accordance  ;

with 6.9.1.2, is an evaluation of the rod worth measurements performed during i startup.

Please address any questions regarding this' submittal to this office.

Respectfully, bh  %

Allen R. Checca-Nuclear Licensing = Administrator cc: :- NRC; Document 1Contro1L Desk-R.-M. Pulsifer, NRR Resident Inspector - Braidwood H. Shafer, RIII 1mw/ID179 N

b 9

4-i --

.' l I

i i

COMMONWEALTH EDISON COMPANY BRAIDWOOD GENERATING STATION b

DOCKET NO. 50-457 LICENSE NO. NPF-77 UNIT 1 CYCLE 3 STARTUP TEST REPORT JULY 1991 PREPARED BYt Jeffrey W. Gurley APPROVED BY:

Javier F. Sanches l

l 262(081591)/1 .

ZD79G s

w -- - .a- , -m +-

IABLE OF CONTENTS CECIlCil PAGE TABLE OF CONTENTS 2

3.0 INTRODUCTION

3 2.0 CORE CONFIGURATION 4 2.1 UNIT 1, CYCLE 2 INFORMATION 4 2.2 UNIT 1, CYCLE 3 INFORMATION 4 3.0 CONTROL ROD DRIVE FYSTEM 8 4.0 INITIAL CRITICALIT" 11 5.0 ZERO POWER PHYMCS TESTING 12 5.1 BORON ENDPOINT MEASUREMENTS 13 5.2 ISOTHERMAL / MODERATOR TEMPERATURE COEFFCIENTS 14 5.3 ROD AND BORON WORTH MEASUREMENTS 15 6.0 AT POWER PHYSICS TESTING 22 7.0 REACTIVITY ANOMALY / BORON FOLLOW 23

8.0 CONCLUSION

AND

SUMMARY

25

9.0 REFERENCES

28 APPENDIX A 29 262(081591)/2 ZD79G W. .

1 A 1.0 11LTRODUCT10H Draidwood Unit 1 Cycle 2 concluded commercial operation on December 30, 1990 after a trip due to a ground fault on the windings of the generator. The Cycle 2 burnup was 13,107 MWD /MTU. This was less then the allowable burnup band for the planned core design. It was necessary to do a core redesign based on 13,107 MWD /MTU.

A total core offload was conducted from March 1, 1991 through March 12, 1991. On March 9, 1991 fuel assembly D29T was found to have a damaged grid strap. This resulted in discharging this assembly and performing a second core redesign. The core was loaded in accordance with the core loading pattern supplied by Nuclear Fuel Services. This core loading pattern was based on the second core redesign.

Unit 1 cycle 3 initital criticality was achieved on May 8, 1991. ~ero power phytics testing was performed from May 8, 1991 through May 14, 1991. Power Ascemslon testing was performed from May 14, 1991 through June 7, 1991.

r r

i 262(081591)/3-ZD79G

e

• .s 2.0 CQ1tE cot {ElGURATION 2.1 Unit 1 Cycle 2 Unit 1 Cycle 2 began commercial operation on December ib, 1969. Cycle 2 concluded operation on December 30, 1990 after a reactor trip due to a-ground fault on the windings of the generator. The cycle 2 burnup was 13.1.07 MWD /MTU. This was less than the allowable burnup band for the plav 96 cycle 3 core design. Table 2.2 shows details of the core regions anc 20mparison of cycle 2 and cycle 3.

2.2 Unit 1 Cycle 3 Damaged Fuel Assemblies During the core of ficed three fuel assemblies were found to be damaged.

Fuel assembly D29T had a bent grid stcap. The decision was made to discharge the assembly. D29T was scheduled for reload in cycle 3. As a result, a second core redesign var done. The current core design is based on this second core redesign. .

Fuel assembly C22S was found to be missing a 1"x1/2" section of the lower inconel grid strap. D07S was found_to have two bent vanes on the corner of the top inconel grid strap. Both assemblies were scheduled for-reload in cycle 3. Utilising a Westinghouse recommendation,-based on industry experience with damaged grid straps and engineering evaluation and safety analysis consistent with 10 CFR 50.59 was provided by Nuclear Fuel Services to allow the reload of fuel assemblies C22S and 007S into the' "ycle 3 core.

I wuclear Fuel Services was also requested to provide a loose part evaluation and safety analysis consistent with 10:CFR 50.59 to allow 1 operating cycle 3 with the missing section of grid strap-from fuel assembly C22S.

Fue. Loading The Unit 1 Cycle 3 core configuration--.is-Illustrated in figure'2.1 The cycle 3 fuel regions 3, 4A, 4B and 4C are Westinghouse _17x17 Optimized Fuel Assemblies (OFA) and fue1~ regions-5A and 5B are Westinghouse 17x17 V ANTAGE - 5.

VANTAGE $ incorporates six design' changes from the existing OFA fuelt

1. - Removable top nozzles for reconstitution.

2. - Additional- plenum space .to allow extended 'burnup.

3.. Natural uranium blankets at the top and bottom 'of . each fuell rod to improve neutron' economy.

4. Intermediate flow Mixer- (IFM) grids to improve DNB margin.

5. Integral Fuel Burnable Absorbers (IFBA)- to eliminate' residual burnable absorber penalty.'

6. Debris filter bottom nousles.:

262(081591)/4 ZD790 V .

?

The core design characteristics and the fuel rod specifications for the cycle 3 regions are provided in the following tables.

Table 2.1 Region Enrichtnent w/o U235 3 3.105 4A 4.012 4B 3.803 4C 3.593 SA Non-IFBA 4.000 SA IFBA 3.994 SB Non-IFBA 3.800 SB IFBA 3.802 Axial Blanket 0.723 Table 2.2 Cycle 2 Cycle 3 Regica Loading Number of Loading Number of (MTV) Assemblies (WIU ) Assemblies 2 17.436 41 ------- --

3 27.202 64 19.126 45 4A 6.812 16 6.812 16 4B 10.195 24 B.496 20 i

4C 20.427 48 20.427 48 SA Non-IFBA ------ --

7.337 24 l SA IFBA - - - - - - --

1.981 l SA Blanket ------ -- 0.864 5B Non-IFBA ------ --

10.396 40 5B IFBA ------ --

4.586 5B Blanket ------ --

1.439 Total 82.072 193 82.004 193 Sources and Burnable Absorbers The Cycle 3 core will contain secondary-source rods and burnable. absorber rods. Two. secondary source spiders, each containing feur source rods will be positioned in core locations G-2 and G-14. A total of-768 fresh wet annular burnable absorber (WABA) rods and 4464 fresh integral fuel burnable absorber (IFBA) rods will be included in the core. The burnable absorber rods are used to maintain a negative MTC and to flatten and refine axial power distribution. The location of the sources and burnable absorbers are shown in figure 2.2.

262(081591)/5 ZD79G

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

s 6 ,4

~

BRAIDWOOD UNIT 1 CYCLE 3 C0RE INVENTORY MAP A B C D E F G H J K L M N P R D72U D038 Cels D32T Cl38 0068 081U S20 630 000 100 20e0 320 200 D357 Doel D430 E195 E70T 0700 E71T E248 Deou 0148 D26T jg BeO R10 270 RSS 08520 R25 450 R03 370 Roe 1000 D17T E16S E078 E577 C508 D61U E008 D61U C648 Elti E00$ Ells D34T ag g 1180 1380 24P7De ROI 30 R13 tentw R31 17D M51 24P81% 1370 930 sw D158 E006 D57U C368 025T C118 E87T C338 D23T C373 D830 D028 R45 P4PS3W R48 2630 1110 110 R20 1320 1130 2750 R47 E038 24P68% R34 12 082U 0500 E63T C20s ceOS E60T Desu C038 0600 E61T C528 C498 E6IT 0500 D870 jj 1310 SD R24 130 1080 rep 50% 540 100 700 P4P73% 100 310 R29 212D 1300 0168 E178 C238 D37T E377 0710 E3ef D75U E477 D700 E337 D38T C40s E238 0046 1200 R503 290 730 24Pe1% RS2 P4PS2w R10 t4Pe6% R04 P4 Peen 800 240 RID 400 10 C125 1e0 E59T 440 De7U R12 C305 300 D800 E41T C558 330 74Pe8% 12D 24P97% 1100 E3ef C218 E44T P4P90W 044U 2230 C4es 740 D480 R32 E64T 530 C018 EA g

D21T 2500 D790 Roe E12S r 4P74%

E63T R40 C138 70 0800 R11 E43T P4PO4%

C518 R15 EdeT P4P70%

D74U R20 Clos 500 EteT R30 E018 R4P75%

0500 R17 D30T 23D g

C098 te5T D62U C5es D590 E45T C278 E387 C10s ES2T D55U C228 Delu E66T CO28 350 520 R30 2050 570 300 24P57M 2150 24Peew D4P97% 700 2250 R23 2200 040 y 0118 E228 CO5S DIST E34T D48U E40T D70U EdeT D400 E42T D3ef Ctes E218 D129 1260 R36 150 1170 P4P77% R43 24P05% R37 P4Pe8% 1443%

R40 51D 300 R26 122D 0410 063U E72T Coos C648 E40T D54U Cett Dl3u E35T C448 Ce25 E54T Desu D730 500 20 R16 250 2130 f4Peew 820 050 SED P4F70% eD 280 906 22D 42D 5

0008 E118 DeeU C448 D28T C638 fesT C078 D27T C578 D64U E106 DIOS R35 24P62% P22 2300 1010 400 R27 10 1040 2000 R38 F4P90W R41 g

D30T Eles E056 EteT C16S D42U E048 Delu C208 ES2T, E028 Elas D31T 620 1350 P4P70W R21 140 R33 24P72w R07 2eD M60 ' r4P50w 1330 970 ~

040T 0058 Dedu E208 E60T Desu E56T ElBS D52U D078 033T sed R05 40 R46 OSSID R02 870 R44 90 ' R14 94 2

077U D018 C668 D24T C258 DISS D47U 640 340 210 720 990 121D 81D j

I FIGURE 2.1 ,

262(081591)/6 ZD79G

_ - . _ . _ _ . _ _ . _ . , - _ _ , _ , . . . _ . . . _ _~. .._..._,_ 2 . __.a.._ _ _ . - . - - - , ._

, ,a BRAIDWOOD UNIT 1 CYCLE 3 CORE INVENTORY HAP J N O F E D C 8 A R P N M L K 100 1

S S- 2 1281 1561 156: 1281

-5 80! R4M 156I 24W 154I 24N 80!

4 156I 24N 24N 5

24M 24N 1561 1561 6

1281 24W 24M 24W 24N 128I 7

24N 24M 24M 1561 1561 270 90 8 24N 1561 24W 24W 1561 24N 9

154I 24W 24W 24N 156I 10 1281 24N 24W 24N 24N 128I 11 156I 24N 24W 154I 12 24N 156! 24N 33 801 24H 156I 24N 156I 24N 881

=

S$t 14 1281 156I 1561 1281 15 0

j

$$ SECONDARY SOURCE: t PREVIOUSLY IRRADIATED XI FRESH ASSEM8LY CONTAINING X IFBA RODLETS XM FRESH BA CLUSTER CONTAINING X RODLETS FIGURE 2.2 LOCATION OF SOURCES AND BURNABLE ABSORBER RODS IN -BRAIDWOOD 1 CYCLE 3.

262(081591)/7

> ZD79G 0

3.0 CONTROL ROD DRIVE SYSTEM All control banks wera repositioned from 228 steps to 231 steps for top of travel. This was done to minimise rod wear and is not intended to be permanent.

The control rod drive system was tested and verified to be operational prior to the Cycle 3 startup. Tests performed included bank overlap, eleve cycler times and rod drop times.

The bank overlap checkout proved the ability of the rod drive overlap unit to step control rod banks in a predetermined sequence such that when the control banks are moving, the next consecutive bank is moving and differs by 125 steps. This was also changed from 113 steps due to the rod repositioning.

The slave cycler timing verli' led the proper sequencing of the conttol rod coil currents during the rod withdrawal and insertion sequence. Table 3.1 shows the measured coil voltages and the espected values. All values were within the acceptable limits.

The rod drop times verified that under hot, full flow conditions, all rods will enter the dashpot region in less than 2.7 seconds after the loss of stationary gripper voltage. This is required by Technical npecification 3.1.3.4. The average rod drop time was 1.428 seconds. Two control rods fell outside of the two sigma limit. These two rods were dropped an additional three times and all times were under the technical spacitication limit of 2.7 seconds. All of the rod drops performed were from the new top of travel position of 231 steps. Table 3.2 provides a summary of all the rod drop times.

Ts. step counters, annunciators, digital rod position indication system an ulce to analog (P/A) converter were all observed for proper ope ability.

262(081591)/8 ZD79G l . . . . . . . . . . . . . . . . . . . . . . . . . .

j

TABLE 3.1  :

+

Slave Cveler Coil Voltaaes and Rod Sceed y l

l Rod Cabinet Direc on Lift'(Volts) Movable Stationary (Volts) Rod Speed *

1. Full Reduced (Volts) Full Reduced (Steps / Min)

H6 1AC. 'Out: 0.57 0.24 0.54 0.51 0.27 49.6 l In 0.58 0.24 0.53 0.51 0.27 50.0 i B4 2AC . Out' O.52 0.20 0.53 0.52 0.27 60.1

' In ' O.54 0.20 0.53 0.52 0.27 64.5 1

g

.G3- '1BD Out. 0.53 0.20 0.53 0.51 0.27 64.2 In 0.54 0.20 0.53 0.51. 0.27 64.5 f*

C7 2BD Out- 0.52 0.20 0.53 0.51 0.27 64.5 '

4 In 0.54 0.20 0.53 0.51 0.27 64.5

)

54 SCDE Out 0.52 0.20 0.54 0.52 0.28 61.5

.In 0.54 0.20 :0.54 0.52 0.28 N/A i

i

Espected Coil Voltacfes (volts)

!- Lift Full - 0.438-0.600 Lift Reduced - 0.163-0.240 2 Movable - 0.438-0.600 l Statieaary Full - 0.438-0.600 i Stationary Reduced - 0.238-0.'330 i; Espected Rod Speed (Steps / Min) i

~

Shutdown Banks'64 1 7.2

~ Control Banks 48 7.2' r t

s 4

{~ ~* Out direction rod. speed is based upon a stopwatch measurement. In direction rod speed is read from 2SI-0412 either/or a stopwatch measurement.-  !

4.. ,

262(081591)ZD79G/9 k'

n >

TAllLE 3.2 HCCA Ditor 71 Hts RCCA Dashpot Entry RCCA Dashput Entry istconds) (s e c9ais.)

D02 1.435 HOB 1.415 111 2 1.415 110 6 1.410 H14 1.490 111 0 1.430 PO4 1.455 708 1.425 B04 1.385 K08 1.390 D14 1.430 702 1.365 P12 A.365 D10 1.515 H02 1.430 K14 1.460 G03 1.430 '06 1.405 C09 1.440 h6 1.445 J13 1.390 ri4 1.415 N07 1.440 P10 1.365 C07 1.425 K02 1.455 013 1.440 180 2 1.425 N09 1.415 B08 1.430 J03 1.465 1114 1.420 E03 1.485 P08 1.420 C11 1.420 F06 1.440 L13 1.520 T10 1.425 NOS 1.405 K10 1.430 C05 1.440 K06 1.410 E13 1.440 D04 1.440 N11 1.410 H12 1.430 LO3 1.410 D12 1.430 1104 1.390 H04 1.375 D08 1.420 110 8 1.470 1112 1.430 262(081591)/10 ZD790

. . . -_-____a

- _ - - - . _ - = . - - . -_-.- -.. . -._ ._ _ ..-.....-..

4.0 11tlH ALCRIIICALITY The inillal control rod withdrawal began on May 7, 1991 at 1537 hours0.0178 days <br />0.427 hours <br />0.00254 weeks <br />5.848285e-4 months <br />.

Control red withdrawal was stopped with Control Bank D (CBD) at 180 steps with approximately 100 PCM remaining worth, Dilution to criticality began on May 7, 1991 at 1620 hours0.0188 days <br />0.45 hours <br />0.00268 weeks <br />6.1641e-4 months <br />, with CBD at 180 steps and an initial boron concentration of approximately 1738 ppm.

Unit 1 achieved criticality at 0015 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> on May 8, 1991 with CBD at 180 steps and a boron concentration of approximately 1315 ppm. A plot of Inverse count rate ratio was maintained as a function of primary water addition during the dilution to criticality.

l l

l l

I 262(081591)/11 ZD790

, ,e 5.0 EERO._PIMER_rilYSICS_TESIlllD Zero power physics testing was completed on May 13, 1991. All Technical Specifications were met. The Moderator Temperature Coefficient (HTC) was calculated to be positive at the beginning of life. It was determined i that HTC would remain positive until sufficient burnup Aas accumulated. j As a result rod withdrawal limits were established to ensure that the HTC '

would remain negative throughout core life.

The sero power physics tests include:

1. Boron Endpoint Heasurements

2. Isothermal Moderatur Temperature Coefficlept Heasurer.u ts

3. Rod and Boron Worth Measurements

4. Shutdown margin verification The results of these tests are discussed an sectiono 5.4, S.3, 5.3 and 5.4.

1 262(081591)/12 ZD79G

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

o /

< 5.1 DQt00.CDdP91nLMenlutamenta The purpose of this test was to determine the ciltical boron concentration for the following rod configurations.

2. All Rods Out,(ARO)

2. Reference Bank In The reference bank is nortnally the bank with the highest predicted worth when inserted in an unrodded core. Since CDC had the highest predicted worth, it wes usted es the reference bank.

The boron endpoint rest *1ts were as follows:

Condition Meantement.(symi Desig:LAcceptance_Criterio ARO 1342 1359 3 $0 ppm i l

CDCin 1246 1253 3 50 ppm Thea measured boron endpoints were within the design acceptance criteria.

l 262(081591)/13 ZD790

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

. i 5.2 I s olhe.rma LTetwetatu r e_Co ci f i cie n LH e ns ur eme nt s The purpore of this test was to determine the 1sotherrnal Temperature Coefficient (1TC) of reactivity. The ITC was determined by measuring ITC for various heat up and cool down cycles. The value of the doppler temperatute coefficient (-1.70 pcm/*T) was subtracted from the mencered ITC to determine the HTC which was then adjusted to all rras out, hot full power, no menon and critical boron conditions. The results of these renasurements and calculations are shown below in table 5.1.

Table 5.1 Isothermal Temperature Coefficient Measurement CBD ITC (pern/'r) HTC **

Cycle position Hessured Predicted * (pem/*r)

Cooldown 1 211 -1.869 0.204 Ileatup 1 210 -2.250 -0.190

-2.50 .t 2.00 Cooldown 2 211 -1.500 0.572 Ileatup 2 211 -2.630 -0.555 AVE MTC 0.008

• Predictions from reference 1 and design criteria from referones 5.

• • Corrected for rods, temperature and boron concentration.

l Since the calculated MTC was positive, rod withdrawal 11m,8ts were implemented per station procedure. When the core burnup en'eeded 084 effective full power hours (ErPil) the value of MTC became negative.

As a reult, the withdrawal limits were removed. Since MTC becomes more nagative with burnup it is expected to remain negative through the remainder of core life.

262(081591)/14 ZD79G 1140/14

1 .

\

I

. s l

1 5.3 RoLandJiotsn_HoIL1LHessurements i The purpose of this test was to determine the differential and  ;

integral worth of the reference bank over its entire length in an unrodded core and to determine the integral worths of the i i

remaining banks were measured using rod exchange method. i The rod enchange technique required calculations by Nticlear ruel  ;

Services (Reference 2) providing estimated critical positions of the reference bank after exchange with the bank being measured, i hP, and the associated correction factors, a,. The measurements L

were obtained for three reference bank positions a) Initially fully inserted position, (h I

• b) Critical position after exchange, h t

c) Final fully inserted position, (h ) Return Theworthofameasuredbank,yf,las wfa w - (Opg ), **(092}x 4

where  !

1) y = The total measured worth of the reference bank.

2) (Apg ), The reference bank worth from 0 stops to the average of (h I o itial and (h ) Return ,

3)' (Ap I2 n The reference bank worth from bW to 231-steps.

i

-4) *, e A correction factor for the h worth due to-the rodded geometry.

262(081591)/15 ZD790- >

,,,,,,,pm,y .,p_, , , n47,.,,,, ,-..m,., , , , , , y,

5.3 Rod _Northlieasurement_Results (cont'd)

Rod worth measurements were performed under the guldance of Westinghouse Rod Eschange topleal (Reference 3). The document states that the allowable percent dif ference between measured and predicted reference bank worth is 10%, 15% for individual rod banks and 10% for the sum of all measured rod banks. For indivldual banks with a predicted worth less than 667 pcm the allowable difference is 100 pcm rather than 15%.

The worth of reference bank (CDC) was measured by dilution. The percent difference between th9 prodleted worth and the measured worth by dilution of CDC was determined to be -9.5. The results of the rod worth measurements based on the dilution measurement of CDC are shown in Table 5.3.a. Using the rod exchange swap technique, all control and shutdown banks were exchanged with CDC. The test criteria for the measured banks CDD and SDB was not met. The test criterla for the total bank worth was also not met.

Nuclear Fuel services was notified of the discrepancies and Westinghouse Electric Corporation was consulted. A review of the rod swap procedure and a reactivity computer checkout was done to identify errors in measurement equipment or procedurel guidance. None were found.

The reference bank was remeasured and confirmed the inltial measurement. Since the measured value of the reference bank was less than predicted, a special procedure was written to measure the remaining bank worths by the boration/dllution method.

Tahle 5.3., Table 5.3.h., and Figure 5.3.a are comparisons between the results of the rod swaps and the boration/ dilution tests.

Nuclear Fuel Services and Westinghouse Electric Corporation performed a review of the Reload Safety Analysis and Safety Parameter Interation List (RSE/SPIL) for Draldwood Unit 1, Cycle

3. The review confirmed that the RSE/SPIL for the cycle was still valid for the core design. Nuclear Fuel Services recommended performing additional flux maps and were included in the startup schedule and performed at various power levels. An onalte review was performed to review the additional data.

Based upon this review, continued power escension was approved.

The cause, safety significance, and corrective action of the rod worth discrepancy are discussed in Attachment A.

The rod exchange method determines boron worth data from the dilution of the reference bank. The CBC differential boron worth from dilution was measured to be -8.88 pcm/ ppm, while the predicteggvas -B.86 pcm/ ppm (ref. 1), a 0.23% difference. There are no acceptance criteria associated with the boron worth calculation, however, 0.23% is considered good agreement between measured and predicted. l 262(081591)/16 ZD790 1140/16

J TABLE 5.3.A ROD EXCHANGE MEASUREMENTS s

]

DILUTION i- BRAIDWOOD UNIT 1, CTCLE 3 I Reference i Reference Bank Bank Inferred Predicted Percent g Withdrawn Worth Worth Difference BANK Inserted (h,)o(steps) g y h, c (4 1), (Q 2), a,(4 I2n "x

. No. ID Initial Return Average (steps) x (pen) (pcm) (pcm) (pcm) (pcm) (%)

3 1 CBC - - - -- -- -- - --

849.9 939 - 9.5 L .

i 2. CBD 36.5 36 36.3 185 1.22 24.59 226.0 275.7 . 549.6 611 -10.1

! .. [

]L 4

3 CBB 36' 36.5 ' 36.3 : 174.5 1.22 24.59 281.6 343.6 481.7 611 -21.2 i'

' I 4 CBA 35 36 35.5 126.5 0.85 23.46 488.9 415.6 410.8 438 - 6.2

+

5 .SBE :34 35 34.5 137 0.91 22.04 445.8 405.7 422.2 497 -15.1 l j- -6 SBD 32. 34 l 33 143.5 1.14 19.92 418.5 477.1 352.9 418 -15.6 I 7 SBC 32 32' 32- 144 1.14 18.50 416.4 474.7 356.7 418 -14.7 8' SBB 36 3B '37 202.5 1.01 25.58 114.7 115.8 708.5 869 -18.5 I e 9 SBA 30 32 .31 143 1.17 17.25 420.6 492.1 340.6 368 - 7.5 lL i n ,

-Measured Integral Reference Bank North, Wg'= A19J (pcm) Total 4472.9 5169.0 -13.5 l

I M' I I2r

! W, =WR Calculations: - IM 1) ~ "x M -

I P P Percent Difference = (W, - W,)/W, x100 Note: ." Predicted values from Reference 2.

1.. '262(081591)/17- -

ZD79G

. .o Table 5.3.D ROD WORTH MEASUREMENT CCHPARISON BANK ROD SWAP DILUTION BORAT10N AVERAGE ERROR (pem) (pcm) (gem) fPCmL_ (\1 CBA 410.8 420.3 426.5 423.4 3.1 CBD 481.7 507.5 454.6 481.0 0.1 CBC 849.9 839.7 842.0 840.8 -1.1 CBD 549.6 572.2 516.5 544.4 -0.9 SBB 708.5 711.3 707.1 709.2 0.1 TOTAL 3001 3051 2947 2999 -0.1 Notes SBB is the highest worth shutdown bank.

CBB data was taken with CBD not fully withdrawn.

262(081591)/18 I ZD79G

0 Til SUF met S TIG. 5.3.a.

1000

---J 900 SW 700 z

8 600 r 500 3 i =

300 -

l- -

?

j -

g l

g

/

d' -

c  ; -

l a j g y ll CBA CBB CBC CBD BA SBB SBC- SBD SBE ROD SWAP MEASURED' [M PREDICTED DILUTION b BORATION-I 2 081591)/19

4 8 6.0 ALrfMILP11YSICLIESIING On May 18, 1991 at 1600 hours0.0185 days <br />0.444 hours <br />0.00265 weeks <br />6.088e-4 months <br />. Unit I was synchrontred to t he gr id for t he start of Cycle 3. Several full core flux maps were taken and analysed to support the power ascension program. The results of the at-power flus map measurements are summarleed in the following sections. All safety and design acceptance cri*eria were met.

6.1 r19x_ Hop _fswe tDisttihutio:LHensutements power distribution measurement.s were performed to support the power ascension. The flux map results showed no indications of a misloaded assembly. All peaking factor limits were met.

The results of the flux maps are summarized in Tables 6.1.A and 6 . 1 . 11 Table 6.1.A shows the reactor conditions at the time of each flux map. Table 6.1.B shows the measured peaking factors, the peaking factor limits, and the power quadrant tilts for each map.

262(081591)/20 ZD790

I l

. L TABLE 6.1.A ,

i FLUX MAP CHARAC M ISTICS .

I Map Date Power Level CBD Position Incore Axial Burnup Ntr.abe r  % (Steps Withdrawn) Offset (%) (MWJ/ CJ) l l

BW10301 05/14/91 7.0 a -18.108 0.0 i i

BW10302 05/21/91 28.6 192 10.450 37.B BW10303 05/24/91 49.1 207 8.561 94.8 BW10304 05/25/91 48.4 180 -6.961 106.6 t I

BW10 05 05/25/91 49.1 207 25.145 115.9 big 3306 05/31/91 73.4 214 -6.330 255.6 BW10307 06/04/91 96.9 215 -0.477 393.9  !

BW3 ,3 0 8 06/04/91 96.9 204 -7.665 409.0 BW10309 06/04/91 97.1 202 2.598 422.4 1

1 2

i I

f I

262(081591)/21 ZD79G ,

i TABLE 6.1.B Fluz Map Results -

Measured Peaking Factors Peaking Factor Limits Incore Quadrant Tilts i

FNDH FNDH Map OFA V5 Fq OFA V5 Fq 41 42 43 44 Number ,

f BW10302 1.5524 1.5507 2.0755 1.8820 2.0034 4.6750 0.9983 1.0121 1.0053 0.9842 BW10303 1.4959 1.5069 1.9682 1.7867 1.9020 4.7375 0.9971 1.0108 1.0053 0.9868 BW10306 1.4759 1.4857 2.0085 1.0737 1.7817 3.1761 0.9974 1.0085 1.0045 0.9869 BW10307 1.4730 1.4708 1.7196 1.5644 1.6653 2.4703 0.9987 1.0071 1.0046 0.9895 262(081591)/22 ZD79G .

e 7.0 EEACI1Y1Tf_AtiOHALUDRE0tLLOLLOW ,

A Core Reactivity Balance was performed per station procedure on June 21, 1991 at a burnup of 1049.9 HWD/Mn). The results of this reactivity balance indicated a difference of 42.9 pcm between predicted and measured values. This difference was well within the acceptable difference of 1000 pcm.

The Core Reactivity Balance Normalisation was performed per station procedure on July 25, 1991 as required by Technical Specification 4.1.1.1.2. The critical boron concentration for all rods out, 100% power, calculated by TOLLOW was 879 ppm. The predicted critical boron concentration was 880 ppm. The measured value differed from the predicted by 1.0 ppm. Because the difference was less than criteria of 5 ppm BwCB-1, rigure 11 was not revised to reflect this blas.

Figure 7.1 shows a plot of measured and normallsed predicted critical boron concentrations verses burnup. The measured values are within the requirement of ,t 1000 pcm as stated in Technical Specification 4.1.1.1.2.

262(081591)/23 ZD79G

, o BRAIDWOOD UNIT 1 CYCLE 3 BORCet TOLLOW DATA TIG. 7.1 1300  :-- -- -  :-*- *- - - - - - - - ~~* - - -- -

1200 1100 -

1000 ,

>+e+ ,

900 A2 ww-

==

o . .

y 000 - -- .H --- - - - - . . - - ... - - ---- -.. -- -. - --

H C 700 -

s .. . - - - . ~ . .. . . .

O '

N s .

C <i

. 600 , .

4 p 500 ,,

400

's li 300 - -_

200 e #

100 -

i i.

80 0 ' *

-*;" -~ ;*

.. . . . . , . .. .. .. . .- . T- . T-o i 2 3 4 s- a 7 a e i .1 i i i i i i i 0 0 0 0 0 0 0- 0 0 -0 1 2 -3 4 6 f. 7 6 0 0 0 0 0 0 0 0 0 0 0 'O O O O O O -O' O O O O O O O O O O O O O O O O O O-0 0 0 0 0 0 0 0 0 RURNUP (WWD/W1U) 262(081591)/24 ZD79G

8.0 SVHMARLRID_CCllCLUSIEtfS The Startup Testing program was completed on June 7, 1991. The results of the testing program as supplemented by the conclusion teached in Attachment A, are considered satisfactory. Table 8.1 provides a summary of the results obtained during startup physics testing Braidwood Unit 1, Cycle 3. Table 8.2 provides a summary of design acceptance criteria and safety acceptance criteria from Ret'tresce 5. All Technical Specifications have been met, acceptance criteria descrepancies resolved, and corrective  ;

actions implemented. Therefore, the startup testing program is considered j to have verified the design analyses for Braidwood Unit 1, Cycle 3.

-262(081591)/25 ZD79G

8.O SUMKARY_AND._CORCLUS10N (Continuedj TABLE 8.1 PHYSICS TESTING

SUMMARY

.. l EARAMETEILMEAEURED _MEAEURED- _ IREDIC2ED _ .nlIIIRENCIL_*

INTEGRAL ROD WORTH MEASUREMLKIS_lpsmL . CBA_ 410.tB (18 - b.1% -

_CDB 411 7 .. 611 -21tD

_ _CBC . 8th.9 939 - 9 th%

._CBD_ 549 6 .611 6-A

. -. ERA - 340.6 11L._ _ .Jt %.-

. _sBB. 708.5 369- _ __.J h 1

_ sac 3$n.1 11a_,,__'__ ,,,r.16 7 %

-_E8D_ 352.9 (1)__ -15.6%

.._sar 4214_". 497 . -lbl% _ , , ,

2QIAL (127 $169 -13.h ISOTHERMAL TEMPERA'n!RE

.,__C0KErlCIERI_fpem/*r) 2 . 0_6 -2.$a 0 14 MODERATOR TEMPERATURE

_CQ%EllC}EKT_(pcm/'T1 0.008 -0.8A2 fl . Big D7frERENTIt.L BORON WORTH.ARO-(pcm/ ppm)

-8.BB -1 36 0.2%

ALL RODS OUT. HOT ZERO POWER CRITICAL BORON CONCENTRATION-(pam) .. 1342 13$9- --7 one CBC IN-NOT EERO POWER CRITICAL BORM CONCENTRATICEl (nema) .1116 '12.$ 3 - 7 ann ALL. RODS OUT. HOT TULL POWER 50R00t-

. . CWCEHIRATICBf (pnel 879' 880- - 1 enn

• See Table 8.2 for the acceptance criteria

• • Average values are given when parameter was measured more than once..

262(081591)/26 .,

ZD79G L-

, 8.0 SUMMARLMD_EWCL951WS_1Continuedl TABLE 8.2 ACCEPTANCE CRITERI A Test Parameter Acceptance Criteria

1. HZP Critical Boron (Control Rods Withdre,- 3 3 50 ppm HZP Critical Doron (Ref erence Bank !# s i i4* 1 50 ppm

2. Differential Boron Worth (Boron Reactivity Coefficient) 3 15%*

3. Control Rod Worth Individual Bank 3 15% or t 100 pcm Review or Design Criteria whichever is greater (for tod swap. the reference bank should be within 1 10%)

Sum of Banks 3 10%

Review or Design Critella

4. ITC 1 2.0 pcm/'r MTC < 0 pcm/*r for entire cycle

5. Plus Symmetry - Incore rius Hess'irement (1 30% RTP).

a. Usirg the Edit for Symmetric thlmbles. 11 0.1 (10%)

Ratio of highest to lowest normalised reaction rate integral for each set of symmetric thimbles.

b. Using the Final Comparison of Saturation 3 10%

Activity and PDQ Activation. Regionwise (For unrodded percent differences between meas., calc. core planes reaction rate integrals for thimbles with only)

> 0.9 RPD **.

c. Using the Symmetry Chec't and Assembly FDHN < 1.02 quadrant power tilts.

6. Power Distribution.

a. Using the Measured and Predicted r gedit. 1 0.1 RPD The differences between the measured and predicted TDHN for each measured assembly,

b. Using INCORE edit " Difference in React. < 0.05 Rate Integrals" value J abeled "Std. Deviation".

7. Critical Boron (100% Power) 1 50 ppm NOTE: Table is based on Amer:4can National Standard ANSI /ANS 19.6.1 - 1985

• NOTE: For calculating percent differences use (Erid - 1) x 100%,

Mens except when calculating percent differences for rod worths use (Mang -1) x 100%

Pred

• • RPD Relative Power Densit:f = Normalized Reaction Rate Integrals or r{g 262(081591)/27 ZD790

_ _ _ _ ~ _ _ _ _ _ __ _ _ _ . _ _ _ _ . _ . _. _ _ _ _ _ _ _ _ __ _ ___ _ _ __ _ _

9.0 REf1RIl{CES

1. NFSR-0089, "duelear Design Report for Braidwood Unit 1, Cycle 3",

April, 1991.

2. CECO Letter, V. S. Noonan to D. L. Tarrar, "Eyron/Braidwood kod Swap Technique," September 9, 1986.

3. CECO Letter BRIC3/042, R. Chin to D. E. O'Brien, April 4, 1991.

4. WCAP-9863-A, " Rod Bank Worth Measurements Utilizing Bank Exchange",

datea Hay, 19t2.

5. FM7/N45e19.6.1-1985, "American National Standards Reload Startup Thy

• hp Dat for Pressurised Water Heactors," December 13, 1985.

e. " h tage 5 Reload Transition Safety Report for the Byron /Braidwood Stations Unita 1 and 2, " Westinghouse Electric Corporation, July, 1990.

7. " Final iteload Safety Evaluation Transmital," Braldwood Unit 1 Cycle 3.

l 262(081591)/28 ZD79G.

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

A

$ 4 APPENDIX A l

262(081591)/29 ZD790 .

August 16, 1991 Evaluation of Braldwood Unit 1 Cycle 3 Rod Horth Hea:Urements ZNLD/1139/1

Table of Contents Section I: Background / Introduction Section II: Description of Event Section III: Results of Detailed Review of the BHIC3 Core Design Section IV: Results of Detailed Review of the BHIC3 Heasurement Methodology Section V: Corrected Test Results Section VI: Conclusions and Corrective Actions t

Section VII: Safety Significance i

ZNLD/1139/2 i

I. Background / Introduction The Braidwood Unit 1 Cycle 3 (DHIC3) reload core is a standard " Low Leakage" design and is similar to the other current Byron and Braidwood cores (BYlC4, BY2C3, and CH2C2). The reload design was initially performed by Westinghouse. Subsequently, a redesign effort was initiated due to the impact of an unscheduled outage, caused by a fault in the unit's electrical generator, which invalidated the reload design's burnup assumptions. The redesign was performed by Commonwealth Edison't Nuclear Fuel Services department and considered the impacts of the premature discharge of fuel assembly D29T and the reconstitution of fuel assembly 072U. Fuel assembly D2" aM three other symmetric assemblies were discharged followinc tha fuel Inspection and reconstitution program when it was discoveref '.nat Vuel assembly D29T sustained damage to the #1 (lower inconel) grid strap during refueling operations.

Braidwood 1 Cycle 3 is a " transition" core and contains a mixture of fresh VANTAGE 5 assemblies and previously loaded Westinghouse OFA assemblies. The NRC approved the use of VANTAGE 5 at Braldwood Unit I for Cycle 3 and thereafter,- The Braidwood UFSAR and safety analysis presently reflects this transition to VANTAGE 5 fuel.

In 1983, Commonwealth Edison Company (CECO) submitted to the NRC a topical report on the benchmark of its nuclear design methods for its PHRs. That report was reviewed and approved. Since that time, CECO has completed over 20 reload nuclear design cycle analyses for its Braidwood, Byron, and Zion nuclear plants. These analyses include loading pattern determinations, verification of neutronic parameters used in Safety Analysis, and generation of nuclear data required for plant startups, operations, and surveillances.

The Ceco methodology uses Westinghouse Electric Corporation arograms_

which are NRC approved for nuclear design analysis and have seen used extensively. Hestinghouse's present core design methodology uses transport theory cross sections and an advanced nodal calculational method. CECO recently modified its design methodology to incorporate these present generation Westinghouse methods, PHOENIX-P and ANC. This transition to the advanced nodal methods was approved by the NRC add applied in-the redesign of the Braidwood 1, Cycle 3 reload core.

It is Commonwealth Edison's practice to verify the adequacy of all relo.'I designs through reload physics testing per the guidelines and recommendations of ANSI /ANS 19.6.1-1985, which specifies the content of an acceptable PHR startup physics test program. The test acceptance criteria _are predetermined values based on industry experience. The test criteria are generally not based on assumptions made in the safety analysis or technical specifications. If an unexpected result is outside' the acceptance criteria of the test but within the conservatisms and limitations assumed in the safety analysis, then testing and power ,

escalation may proceed. A specific individual test is evalucted th conjunction with the overall results obtained during the startup test-program to demonstrate the adequacy of the reload-design. -Successful l completion of the physics test program provides assur_ance that_the I reactor can be operated as designed.- -

ZNLD/1139/3

F

. 4 The CECO reload physics test sequence includes a test to measure the rodworth of all control and shutdown banks. CECO performs the measurement using one of two methods, the " rod swap" or "boration/dllution". It should be noted that CECO has succes' fully performed rodworth measurements after each Zion Byron or BraJdwood ,

reload (approximately 35 measurements) without any significart i discrepancies.

In summary, no aspect of this reload Ossign or associated physics testing effort was a significant departure from past CECO or Westinghouse practice or range of experience.

II. Description of Event Braidwood Station began )hysics testing on May 7, 1991, according to BwVS 500-3, " Reload Startup P1ysics Tests following Refueling." The point of adding nuclear heat was found and the determination of the low power physics range was made. Later, BwVS 500-5, Rod and Boron Horth Heasurements, was entered to determine the integral rod worth of all banks using the rod swap technique. The integral worth of the rod banks was determined to be 4472 pcm, as compared to the predicted 5169 pcm. At this point, the testing was halted to investigate the difference between the measured versus nredicted rod worths. Since the BHIC3 predictions were consistent with previous Byron and Braldwood rod worth predictions, while the measured value significantly deviated from this data, initial indications were that the measured values were suspect. The reactivity computer, the testing procedure, the rod swap prediction and its supporting analysis, and the startup data developed by CECO's nuclear designers were evaluated ar;d no discrepancies were identified. At that point, Commonwealth Edison re-performed the measurement of the reference bank. The re-measurement gave the same results as the original measurement. Since the worth of the reference bank was less than 3redicted, it was decided to measure the worth of all banks using the

) oration /dtlution method. The boration/ dilution method is considered to be an independent and accurate measurement technique. The addltional rod worth measurements were performed for all control banks and the highest worth shutdown bank, and essentially reproduced the previous rod swap results.

Ceco performed a re-evaluation of the neutronic portion of the BHIC3 Reload Safety Evaluation taking into account the measured rod worths.

The re-analysis demonstrated that the reload core's neutronic analysis, as modified by the rod worth results, remained within the bounding assumptions and conclusions of the Byron /Braldwood UFSAR safety analysis.

Following the re-evaluation, braidwood Station perfor'wd a controlled power accension adding additional holds-for flux maps as recommended by.

the CECO Nuclear Fuel Services department. Flux map measurements were taken and evaluated at 8% power, 30% power, 50% power, 75% power, 90% power, and 100% power. The flux map results at each plateau:-

- confirmed that the core was operating within technical specification and safety analysis limits, ZNLD/1139/4

_ ._ ~._ _ _--

- confirmed that the predicted power distributtons and the core's measured power distributions were equivalent, and

- confirmed that there was no core misloating or other physical anomalles.

However, in order to determine the cause(s) of the testing discrepancy Ceco, in conjunction with its vendor Westinghouse, initiated a comprehensive " ground up" evaluation of both the BWIC3 core model and the PHOENIX /ANC design methods.

III. Resu11s_of_DatalleLReview_oLihe_RilC3 core Datino The BHlC3 core model was generated with the Westinghouse ALPHA / PHOENIX /ANC code package, an updated and automated version of the l PHIX/ PHOENIX /ANC code package used in CECO's benchmark evaluation of the i PHOENIX /ANC methodology. The benchmark effort showed very good agreement between code predictions and historical measurements, while the-initial i

, application of the automated ALPHA version (BW1C3) had a predicted versus measurement difference. The possibility that minor code errors were introduced in the transition to the new version was examined, l

Ceco has completed two evaluations using the original PHIX/ PHOENIX /ANC l version as an independent calculation method. Ceco regenerated the BHIC3 model with PHIX/ PHOENIX /ANC. Ceco also regenerated the core model for Byron 2 Cycle 3 (BY2C3) with the new versions of ALPHA / PHOENIX /ANC. The model that was used for the original BY2C3 startup data was generated using the PHIX/ PHOENIX /ANC code package. These two models for each of .

the two cores (bHIC3 and BY2C3) allow a direct comparison between the versions of the codes.

I Rodswap data.was regenerated for both cases. For BHIC3, all physics test acceptance criteria would have been met with the PHIX/ PHOENIX /ANC model.

The PHlX/ PHOENIX /ANC models still overpredict most of the banks, but the total-bank worth error is 6% instead of 13.5%. For BY2C3, both code packages would have satisfied all acceptance criteria for startup physics testing; however, the predictions with ALPHA / PHOENIX /ANC are not as close to measurement results as the original model. .It was later-shown that both code _ packages give excellent results When minor modeling and code errors and corrected measurements were considered.

Upon CECO request, Westinghouse generated an independent core model for BHIC3 for comparison with the design of record. Nestinghouse's independent model validatas.the CECO analysis. Besides generating an independent BHIC3 model, Westinghouse audited CECO's investigation and-reviewed CECO's models and methods. Nestinghouse found that Ceco's and Hestinghouse's methods are generally: consistent but_some slight changes in the method CECO used to generate the control rod cross sections was-  :

recommended. Westinghouse also notified Ceco of two_ minor errors-in their ALPHA / PHOENIX /ANC code _ package-which also affected the CECO version: ,

1. An error in the hafnlum number densities used in the calculation, and' l

2. An error in the unit assembly mesh space.for hafnium control rodL calculations.

ZNLD/1139/5

~

v- , N~- -s t e -Nn e- ,<m--w-ws- ~~ e-

, The improvement from the code corrections and improved modeling is demonstrated below in Table 1:

TABLE 1 DHIC3 R00 HORTH PREDICTIONS Original / Uncorrected Corrected CECO Independent

__ B a nL__. _Al.EHA/EHOENIXLANC__ 6LPJiAlfil0ENIX/ANC PflIX/fBOLNIX/ANC CBC (Ref) 939 897 909 CBD 611 584 572 CB8 611 581 539 CBA 432 424 445 SBE 497 467 459 SBD 418 395 372 SBC 418 395 372 SBB 869 831 806 SBA 369 353 330 Total 5169 4927 4804 CECO and Westinghouse are confident that the investigation of CECO's nuclear design methods and codes described above identified and corrected the problems introduced in the transition to ALPHA / PHOENIX /ANC. This joint effort will result in an improvement in the accuracy of future rodworth predictions; however, the investigation did not adequately resolve the BWIC3 rod wortn issue. A significant discrepancy between measured and (corrected) predictud rod worth remained, CECO and Westinghouse concluded that it was very likely that:

1. Although the BWIC3 rodworth measurements are repeatable, the measurements probably contained a systematic measurement error, and

2. This systematic measurement error (in a weak form) has been causing the traditional under measurement in CECO's rodworth measurements.

In response to CECO's regaest an independent investigation of CECO's rod worth measurement methods and procedures was performed by Westinghouse.

Results are described in Chapt 6r IV of this report.

ZNLD/1139/6 1

a

k IV. Results of Detailed Review of the BMIC3 Rod North Measurements p Since Commonwealth Edison had verified that the rod worth measurements were repeatable, and had verified the adequacy of the BHIC3 design models, CECO and.Hestinghouse focused their investigations on the testing methodology. The investigation focused on one issue that has introduced systematic errors in rod worth measurements at other Westinghouse units:

-the effects of " Gamma" background on rod worth measurements.

Gamma background contamination occurs when the physics test neutron flux 4 range selected for physics testing is too close to the garm.a background flux level and the resulting contribution from gamma to the flux input signal to the reactivity computer becomes significant. The measured reactivity change, and therefore rod worth,-is reduced as the constant gamma background signal, unaffected by the control bank insertion,- ,

reduces the relative change in indicated neutron population. The impact on measured rod worth can be significant; and has adversely affected rodworth measurements at several Hestinghouse units (most recently at Diablo Canyon).

Ceco and Hestinghouse's investigations did identify gamma contamination in the BHIC3 test results. The contribution _from gamma was 6 to 10% of the signal versus the 1 to_2% which is typical. The impact on the  :,S rodworth measurements is significant and explains the discrepancy between the measured and predicted rodworths. The phenomena was identified and confirmed by examining the signal from the source range and intermediate range neutron flux detectors at zero power.

3 The average indicated intermediate range detector signal present just prior to withdrawing the Shutdown Banks to begin the previous cycle (BHIC2) startup was approximately 1E-05 microamps. The average-intermediate range detector signal present just prior to withdrawing the thutdown Banks in BHIC3 was approximately 1.3E-05 microamps, which ,

represents an increase of approximately 30%.-

the Hot Zero Fower (HZP) Test Procedure-currently approved for use.on.

Braidwood Unit 1 specifies that the Physics Testing Flux Range be. defined by:

1. Determining the indicated flux signal level-at which the' Doppler

-Temperature Coefficient reactivity feedback begins to significantly influence the indicated reactivity change value behavior (Point of Nuclear Heat). The upper end of the Physics Test Range is required to be no higher than 1/3 of the indicated flux _ signal at the Point of Nuclear Heat.

2. The procedure requires that the lower end of the Physics Test Range.

be at least 0.5 decades above the indicated flux level corresponding:

to the point at which the calculated reactivity beginstto deviate

-significantly due to gamma contribution.

q ZNLD/1139/7

The rethodology used to determine the low end of the Physics Test Range specified in the HZP Test Procedure does not directly allow for the determination of the fraction of the input signal due to the background gamma level._ The method provided in the procedur only allows an estimate of when the fraction of the indicated f) 11gnal due to gamma contamination begins to significantly influence the reactivity calculations. The HZP Test Procedure does not adequately ensure that the icw end of the Physics Test Range will be established to preclude the fraction of the input flux signal due to background gamma from exceeding 1% of the total throughout the Physics Test Range.

An examination of the data used to establish the testing range shows that for Braidwood 1 Cycle 3 the " point of adding heat" and "the point where gamma contamination begins to predominate the reactivity signal" are separated by less than two decades. It would have been impossible to establish a testing range two decades above the gamma noise region which is what would have been required to ensure that the gamma signal is less than 1% of the measured signal.

As mentioned previously, the contribution to the total signal from gamma background was 6 to 10% of the signal. Nestinghouse has estimated the impact on the reactivity measurement and determined that a correction.

factor of 1.10 to 1.06 is necessary for the BWIC3 measurements in the middle 20% of the testing range (the significant portion of the testing range).

Application of these correction factors to the Reference Bank worth measurement wou)d change the measured value from 849.9 pcm to between 934.9 pcm and 900.0 pcm. The corrected Rod Swap measurements which result from an average 1.08 gamma reactivity correction factor are provided-in Table '.i 4

ZNLD/ll39/8 -

1. 8 .

9.-

IABLfJ Measured Value Corrected Value Bank (ECM) (ECM)

CBC (REF) 849.9 918.0

, SBA 340.6 368.0 SBC 356.7 385.4 SBD 352.9 381.2 SBE 422.2 456.1 CBA 410.8 -443.9 CBB A81.7 520.4 CBD 549.6 593.7 SBB 708.5 765.3 Total 4472.9 4832.0 CECO has also examined the physics test data for establishing th6 physics testing range at Byron Station and identified similar circumst:mces, i.e. the available range did not always provide two decades between the gamma background level and the testing region. Table 3 provides a summary of previous Byron and Braldwood Rod Horth measurement experience (uncorrected for gamma background): ,

labls_3 Cxt11 Measured (PCM) Predicted (PCM) Difference (%)

Byron 1 Cycle 2 4883.7 5293 -7.7%

Byron 1 Cycle 3 4818.0 5162 -6,7%-

Byron 1 Cycle 4 4705.0 4935 - -4.7%

Byron 2 Cycle 2 4757.7 5162 -7.8%

Braidwood 1 Cycle 2 4564.5 4920 -7.21 Braidwood 2 Cycle 2 5012.3 5093 -1.6%

4 ZNLD/1139/9 j 1

i

_ _ _ _ _ _ _ _ m_____________.____._____.--_._____._____.m._ . _ _ _ _ . _ _ _ _ _ _ _ _ _ - _ _ _ _ _ .

~

'V. Corrected Test Results The following table (Table 4) provides the BWIC3 results after both the design predictions and measured values have been corrected to reflect the results of this investigation.

JABLLJ BHIC3 Corrected Prediction versus Corrected Measurement CECO Independent

_Bant_ ContetteLBeasstement CorretteLALMMEH0 MIX /RC MIXLEH0BIX/ANC (pcm) (pcm) (%) (pcm)(%)

CBC (REF) 918 897 (+ 2.3%) 909 (+ 1.0)

CBD 594 584 (+ 1.7%) 572 (+ 3.9%)

CBB 520 581 (-10.5%) 539 (- 3.5%)

CBA 444 424 (+ 4.7%) 445 (- 0.2%)

SBE 456 467 (- 2.4%) 459 (- 0.6%)

SBD 381 395 (- 3.5%) 372 (+ 2.4%)

SBC 385 395 (- 2.5%) 372 (+ 3.5%)

SBB 765 831 (- 7.9%) 806 (- 5.1%)

SBA 368 353 (+ 4.3%) 330 (+11.5%)

Total 4832 4927 (- 1.9%) 4804 (+ 0.6%)

As can be seen from the table above, the final measured versus predicted values are in excellent agreement.

VI. Conclusions and Corrective Actions Tne results of CECO's and Westinghouse's joint review support the conclusions that the rod swap predictions (and associated design model) and the rod swap measurements were developed and performed correctly.

A. CONCLUSIONS:

CECO and Hestinghouse's evaluations to date have confirmed that:

(1) There is no misloading or basic physical problem with the Braidwood 1 Cycle 3 reload core. This is concluded-from the overall acceptability cf the BW1C3 reload physics tests and verified in detail by three dimensional core mapping of the reload core power distributions using the flux map systems at approximately 8%, 30%,

50%, 75%, 90%, and 100% power.

(2) There is no "at pcwer" discrepancy between predicted and measured power distributions. This was determined from the numerous flux maps taken at the various power levels listed above.

(3) There are_no.significant calculational errors in the CECO BH1C3 l t

design model or in the design calculations as demonstrated in the l detailed review described in Section III of this report (i.e. the code packages were applied correctly and CECO design procedures are correct).

I

! ZHLD/1139/10

. .~a (4) Two minor errors in the ALPHA / PHOENIX /ANC updated design code package were discovered as described in Section III of this report. All future calculations will be performed with corrections to these minor errors.

(5) There is no fundamental problem with the Westinghouse ANC/ PHOENIX methodology. This is demonstrated by the results of Hestinghouse's examination and extension of the continuous " benchmarking" database which includes results from the physics testing for all Hestinghouse-fueled plants.

(6) The primary cause of the difference between predicted and measured rod worths was excess gamma noise in the rod worth measurements as described in Section IV of this report.

B. CORRECTIVE ACTIONS The following items have been completed or are proposed as corrective at.tions : .

(1) CECO and Westinghouse formed a joint working group that investigated and thoroughly explained the differences between the versions of ANC/ PHOENIX-(completed action).

(2) CECO will verify the adequacy of the mechanism by which Westinghouse notifies CECO of code differences and informs CECO of Westinghouse's industry experience when that experience indicates an improvement in a code version is appropriate.-

(3) CECO will verify that the physics test range is.not affected by gamma

-background for each PHR startup-using a constant rod worth check.

The basic principle is to introduce a step control rod insertion or withdrawal throughout the testing range and verify that the reactivity change is constant. Excess gamma background would reduce the worth of the insertion / withdrawal in the lower end of-the test range.

-(4) Most importantly, CECO will implement a change to CECO's reactivity computers which will provide direct gamma background. compensation..

The proposed design has been successfully used on similar hardware at other Westinghouse plants.

VII. Safety Significance A. CECO and Westinghouse have concluded that the Braidwood Un'lt-1, Cycle 3. reload core is operating as designed and have confirmed the adequacy of the BWIC3 design model. Specifically,~the "at-power" flux measurements show excellent agreement between predicted and-measured power distributions; one zero power measurement-(rod. worth) indicated a discrepancy which has been investigated, explained, and considered'in all applicable' safety analyses.

ZNLD/ll39/11 l

. ,~

• As is CECO's standard practice, during the BHIC3 reload cores' Reload Safety Evaluation Process, Ceco verifted that the core will perform under current nominal design parameters, Technical Specifications and related bases, and current Technical Specification setpoints such that:

1. Core operating characteristics will be equivalent or less limiting than those previously reviewed and accepted, or

2. For those postulated incidents analyzed and reported in the Braidwood Updated final Safety Analysis Report (UFSAR) which could potentially be affected by fuel reload, reanalysis, or re-evaluations have been performed to demonstrate that the results of the postulated events are within allowable limits.

In summary, the Braldwood 1 Cycle'3 reload core is operating as designed and it was confirmed prior to power ascension that the design satisfies all accident analysis assumptions, further, CECO has confirmed that both the original design rod worths and the original measured rod worths (and all rod worths in between) satisfy accident analysis rod worth assumptions and limits.

ZNLD/1139/12

_ _