ML20082S426
| ML20082S426 | |
| Person / Time | |
|---|---|
| Site: | Callaway  |
| Issue date: | 09/12/1991 |
| From: | Schnell D UNION ELECTRIC CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| TAC-M80863, ULNRC-2484, NUDOCS 9109170136 | |
| Download: ML20082S426 (9) | |
Text
'
n-5, g :f L, l
s Ni<
(A
,: m
[]NIoN Y""?l!!!""*,
h,lECTIU( '
r.
'[] Geptmier 12, 1991 U.S. 11uclear Regulatory Commlacion Document Control Deck Mail Stop P1-137 Washington, D.C.
20055 UL!lRC-2 4 8 4 TAC 11o. 1480863 Gentlement DOCKET 11UMUl:R 50-4 83 CALLAWAY PLAliT Ull10!1 ELECTRIC REPORT Col 4 TROL BAllK REACTIVITY WORTil DETE104111ATIO11 USIllG Tile ROD SWAP TECil!11(UE Referencet
- 1. UL11RC-2 4 2 9, June 28, 1991
(
llRC staff and Union Electric personnel met in person und by telechone on I.ugust 8 and 13, 1991, respectively, to discuse Union Electric's rod swap g
submittal previously tranamitted by Reference 1.
Thin letter providen the additional information requested at those meetings.
As in Reference 1, we requent that your review of this report be completed by January 1, 1992 to support implementation of rod swap during out next retueling outage scheduled for Spring 1992.
If any additional t,
information concerning this matter is needed, pleaco contact us.
Very truly yours,
, ')
/,//?
,/
l
[
Cb/ga vt {<,
/ Donald F.
Schnell
//p~
GAC/dln a_* __
sg e
cci T.
A.
Baxter, Enq.
Shaw, Pittman, Pottn & Tiowbiidge 2300 11. Street, 11. W Wanhington, D.C.
20037 Dr.
J.
O.
Celmak CFA, Inc.
10225-A Flower Hill Way Gaithersbutg, i1D 20079-5334 R.
C.
};nop Chief, Reactor Project Branch 1 U.S.
thiclet,r Regui nt ory Commi nnion Region 111 799 Roonevelt Rond Glen Ellyn, Illinoin 60137 Bruce Bart lott.
Callaway Resident Office U.S.
thici e n t koguintory Commirnton RRif1 Steedman, 111 n r.ou r i 65077 11. D.
I,ynch (2) of fice of flucient Reactor Regulation U.S.
11uclear Peguintory Commincion 1 White Flint, 110 t th, Mnil stop 13E21 11555 Rockville Pike Rockville, 11D 208B2 Manager, Electric Department Minsouri Public Service Comminnion P.O.
Box 360 Jefferoon City, MO 65102
ULNRC-3 4 84
~
RESPONSES TO NRC QUESTIONS ON UE ROD SWAP METHODOLOGY 1.
Clarification of Equations and Test Proceduro Section 2.1 of the UE submittal discusses fundamentals of the rod swap techniquo, including the rolovant bank worth equations.
The equations are repeated below:
W(X)RB@CP = W(RB) - (W(RB) CP-AR03 whero; W(X)pggep = Reference' Bank at the criticalIntegral worth of Bank X with t position W(RB) = Total integral worth of the Reference Bank with no other banks inserted (measured with the dilution technique)
W(RB)CP-ARO = Reference Bank integral worth from the critical position to fully withdrawn Essentially, this equation states that the worth of Bank X, in the presence of the Reference Bank, is equal to the worth of the Reference Bank (in the absence of all other banks),
from fully inserted to the observed critical position.
However,_ duo to the manner in which dilution-modo worth data is - takers (e.g. Reference Bank worth is measured by insorting rather than withdrawing the bank), it is more convenient to express rod swap bank worths as shown above.
Once the integral worth of the Reference. Bank is measured and all critical position data collected and adjusted (see Item 4), the test engineer uses the measured data and the-equation shown above to determino each bank worth.
Furthermore, it is important to emphasize that no adjustment factors are involved, and-the derived worth values are the same as what would be obtained by explicitly measuring each bank worth, with the Reference Bank at-the respectivo critical position, using the dilution technique.
In satisfying the acceptance / review critoria, the measured values are compared against predicted values _for the same exact configuratione In other words, tabulaticus of predicted bank worth versun Reference Bank position are supplied for the tost.
The test engineer simply looks up (using interpolation if necessary) the predicted worth associated with the adjusted measured critical position.
The resulting worth prediction is then compared against the measured worth.
This process is repeated for each bank.
U LNRC-2 4 84 2.
Reference Bank Pre-Swap position Limit Analysin This issue relates to the maximum number of steps the Reference Bank can be withdrawn prior to an exchange maneuver without excessively impacting test results.
In typical dilution-modo testing, boron dilution is terminated and the core allowed to stabilize shortly before the measured bank reaches full insertion.
If dilution is not terminated soon enough, RCS mixing will cause a small reactivity overshoot, which will in turn require a partial insertion of another bank to maintain criticality.
- Ideally, the dilution is terminated such that the bank just reaches full insertion as the core stabilizes.
However, this rarely occurs, and the bank ends up slightly withdrawn.
To provide complete data for the test, the bank is temporarily incerted to measure the remaining worth segment.
In rod swap, since Reference Bank worth is measured using boron dilution, the above scenario usually occurs.
As a result, the Reference Bar:k typically begins at a slightly withdrawn position before each exchange maneuver.
In the Callaway Cycle 4 and 5 benchmarks, initial Reference Bank position, ranged from 11 to 32 steps, which translates into 2.24 to.'. 6 pcm.
However, as described in the submittal, this situstion is properly accounted for by adjusting the raw measured critical positions.
Thus, the critical positions ultimately used in the procedure are equivalent to the positions that would have been observed had the Reference Bank been fully inserted and test condition drift not occurred.
In all physics measurements, various secondary effects (i.e. boron concentration effects, spectral effects, local flux distribution effects, etc.) can potentially impact test results.
Theoretically, if secondary effects did not exist, then an initial Reference Bank position limit would not be needed.
However, such effects do exist and tend to grow with increasing initial Reference Bank positions.
Therefore, it is important to establish a reasonable Reference Bank initial position limit that ensures test results are not significantly impacted.
Union Electric has performed a sensitivity study to determine such a limit.
The study, which is documented in UE calculation NFDC 91-026, clearly shows that an initial Reference Bank position of as much as 50 steps withdrawn will not excessively impact test results.
In the cases analyzed the impact on individual bank worth ranged from
-2.9% to +3.3%.
Since review criteria rather than ucceptance criteria are applied to individual bank worths, this level of potential impact is not considered excessive.
r U LNRC-2 4 84 i
Most important, however, the impact on total bank worth is only +0.6%.
Based on the analysis, total bank worth, which constitutes an acceptance criterion and is of primary concern, is insensitive to initial Reference Bank positions of 50 steps or less.
In order to maintain a conservative position while ensuring adequate. test flexibility, a Reference Bank initial position limit of 40 steps withdrawn will be incorporated in Union l
Electric rod swap test procedures.
3.
Benchmark Data As documented in the submittal (pga. 41-56), Union i
Electric's rod swap method has been benchmarked against explicit rod swap measurements conducted during startup of callaway cycles 4 and 5.
Six banks were measured in Cycle 4, while all nine banks were measured in Cycle 5.
l Thus, a total of fifteen control and shutdown banks in a
~
wide range of core locations have been measured using the UE method.
As seen in the benchmark results, all measurements easily passed _the respective acceptance / review criteria.
i These benchmarks clearly demonstrate the validity of the Union Electric rod swap methodology, as well UE's ability to i
properly implement the method in a production environment.
Also documented in the submittal (pgs. 57-78) are additional startup physics test and core follow comparisons for Callaway Cycles 1 through 5.
These benchmarks span a wide range of fuel type, enrichment, burnable poison, and control rod designs.
Callaway has transitioned through the Standard and Optimized Fuel Assembly designs to Vantage 5, with i
enrichments ranging from 2.3 to 4.4 weight percent.
Along with multiple fuel designs, callaway has used the Standard Burnable Poison Rod Assembly (BPRA), Wet Annular Burnable r
. Absorber (WABA), and most recently, the Integral Fuel Burnable Absorber (IFBA) designs.
Lastly, Hafnium control rods were used in Cycles 1 through 3, and Silver-Indium-Cadmium control rods in later cycles.
As seen in the submittal, Union Electric calculations consistently show excellent agreement with measurements.
These benchmarks demonstrate Union Electric's overall reactor physics analysis capabilities, and provide further evidence of UE's ability to implement rod swap at Callaway in a proper, reliable manner.
l Based on further NRC review of this data, it was established during the August 13 UE/NRC telecon that additional benchmark data is not necessary.
However, to further document the quality of Union Electric's methods, additional benchmark data is provided in Attachments 1 and 2.
l f
I f
ULHRC-2 4 8 4 l presents rod swap benchmark data for cycle 1 of i
a plant similar to callaway.
As soon in the attachment, all banks passed the acceptanco/ review criteria.
Although porcent errors in those additional benchmarks are somowhat higher than in the callaway benchmarks (see submittal), tho i
error trends are consistent with error trends betwoon measuromonts and predictions generated by the other plant's personnel (this data is not included for propriotary reasons). prosents comparisens of Union Electric and vendor bank worth calculations for callaway Cycles 1 through 5.
When reviewing the data, une may note a few largo deviations betkumr UE and vendor predictions, l
particularly for cycle 2 Herover, in most casos tho measured value is a t leac; ar close to the UE valuo as to the vendor value.
4%r vxarigle, for control Bank C (CBC) in Cycle 3, the UE prodius.;.s differs from the vendor prediction by 11.6%, whlie the measured valut differs from the UE prediction by n'.ly 2.7% (soo submittal pg. 60).
4.-
Referenco nank critical Position Adjuutments As discussed in Item 2, Roference Bank critical position adjustments are necessary to compensate for a non-fully inserted initial Roferenco Bank position and test condition drift (if any).
Adjustments "ary depending cn the banP boing measured as well as tF
.nitial Referenco Bank
[
position, but generally are,n the order of 5 steps or loss.
High-worth banks which have a nominally high Reference Bank critical position tend to have large adjustments since the critical position lies in a low-worth region (near the top of the core).
Conversely, low-worth banks tend to have
(
small adjustments for the samo initial conditions since the critical Reference Bank position lies in a high-worth region (near the middle of the core).
Note, however, that although stor adjustments can vary given the same initial conditions, adjustments will be constant in terms of reactivity.
For examplo, if all swaps began with an initial Reference Bank posicion equivalent to 20 pcm, and i
relevant condition drift of 2 pcm occurred during each swap, t
then all adjustments would be equivalent to 22 pcm, although the adjustments in steps would vary for each bank.
L It should be emphasized that Referenco Bank critical
(
position adjustments are made tb'.fough an objectivo, t
l procedural process, and the quaalty of the adjustments is not materially sensitive to initial Reference Bank position.
r t
I r
~,. _
U1JG:C-2404 S.
Itemodial Actions (Default to Doren Dilution)
To clarify Section 4.3 of the submittal, if any rod swap acceptance critoria are not mot, then testing shall automatically default to dilution testing using Union Electric's current procedures and acceptanco criteria.
6.
Callaway cycle 1 Dilution Measurements Dilution-modo bank worth testing conducted during Cycle i startup included control Banks D, C,
B, and A, and shutdown Bankr, E, D,
and C.
Shutdown Banks B and A woro not required to be measured.
Other testing included worth measurements of thc control banks in normal overlap modo, and measurement of the predicted worst stuck rod.
ULNRC-24 84 ADDITIONAL ROD SWAP III NCllMARKS BANK Manp.
Prod.
% Error Criteria OX1 CBD 651.0 690.8
~5.8
+15/-13.6 YES CBC 94.4.0 958.3
-1.5
+15/-14.4 YES CBB 619./
657.4
-5.7
+12.2/-15 YES CBA 406.8 416.1
-2,2
+15/-14.0 YES SBE 492.3 492.6 0.0
+15/-12.9 YES SBD 433.8 461.6
-6.0
+15/-14.6 YES SBC 416.4 456.3
-8.3
+14.9/~15 YES SBB(RS) 9'/1.1 988.4
-1.8
+10/-10 YES SBA 390.2 435.7
-10.4
+10.6/-15 YES TOTAL 5326 9 5557.2
-4.1
+10/-9.6 YES
- Results documented in UE calc NFDC 91-026 i
f U LNRC-2 4 8 4 COMPARISON OF UNION ELECTRIC AND VENIX)R i
DIIDTION-MODE BANK WORTH PREDICTIONS i
Cycle 1 DARK UE_
VE.HDOR
%DEV CBD 641 650
-1.4 CBC 1232 1240
-0.6 CBB 997 970
+2.8 CBA 673 680
-1.0 SBE 852 870
-2.1 SBD 737 740
-0.4 SBC 961 960
+0.1 SBB 1173 1210
-3.1 SBA 501 480
+4.4 TOTAL 6093 7000
-0.4 Cycle 2 DANK UE VEll.QQR
%Dgy CBD 618 645
-4.2 CBC 1115 1138
-2.0 CBB 921 941
-2.1 CBA 522 543
-3.9 TOTAL 3176 3267
-2.8 t
Cycle 3 DANK UE VENDOR
%DEV l
CBD 538 533
+0.9 l
CBC 872 986
-11.6 CBB 1242 1391
-10.7 CBA 389 384 41 1 2
TOTAL 3041 3294
-7.7 Cyclo 4 BAlfK UE VENDOR
%DEY CBD 701 719
-2.5 CBC 961 937
+2.6 i
CBB 905 861
+5.1 i
CDA 684 642
+ 6.5 t
TOTAL 3251 3159 42.9 Cycle 5 BANK UE VENDOR
%DEV CBD 520 519
+0.2 CBC 1150 1135
+1.3 CBB 1019 936 48.9 CBA 642 656
-2.1 TOTAL 3331 3246
+2.6
%DEV = 100*(UE-VENDOR)/ VENDOR
)
I
-