ML20092H799
| ML20092H799 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 02/14/1992 |
| From: | William Cahill, Woodlan D TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| TXX-92093, NUDOCS 9202210358 | |
| Download: ML20092H799 (33) | |
Text
. _ _ _ _ _ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _.
&J Log # TXX-92093
- llllll" File # 10010
_.~.
915 r
=
7UELECTRIC February 14, 1992 William J.Cahill Jr.
Gnwp We Presi,km U. S. Nuclear Regulatory Commission Atto: Document Control Desk Washington, DC 20555
SUBJECT:
COMANCNE PCAK STEAM ELECTRIC STATION (CPSES)
DOCKET N05. 50-445 AND 50-446 REQUEST FOR ADDITIONAL INFORMATION ON RXE-91-002
- REACTIVITY ANOMAL.Y EVENTS METHODOLOGY" REF:
Letter from the NRC to Mr. William J. Cahill, Jr. dated January 14. 1992, Requesting Additional Information regarding Topical Report RX E - 91 * '102 Gentlemen:
Attached, please find TV Electric's responses to 26 of the 28 questions provided in the referenced letter.
The responses to the remaining two questions require additional analyses and therefore require additional time f or completion.
TV Electric will provide the responses to those questions by March 31, 1992.
Should clarification or additional inf ormation regarding responses to the l.
referenced letter be required to enabic the Staf f to complete its review, contact Mr. Jimmy D. Seawright at 214-812-4375.
Sincerely.
William J. Cahill, Jr.
By:
Id D. R. Woodlan Docket Licensing Manager JDS/grp At t a chinent
.c - Mr. R. D. "artin,-Region IV Resident Inspectors, CPSES (2)
Mr. T. A.
Bergmon, NRR 9202210358 920214 PDR ADOCK 05000445 3gq p
.e)N. ouve street LB. si Dallas, Texas 7520 hl
Attachment to TXX-92093 Page 1 of 32 TOPICAL REPORT RXE-91-002 REACTIVITY ANOMALY EVENTS METHODOLOGY Note:
The references, figures, tables, and nomenclature quoted in this response correspond to those provided in Topical Report RXE-91-002.
1.
Question Have the TUE methods and correlations of References 3-7 been approved for the applications required by the reactivity anomaly methodology of RXE-91-002?
Answer References 3 through 7 of RXE-91 v02 are currently under review by the NRC Staff, and have not yet been approved for the applications required by RXE-91-002.
The intent of RXE-91-002 is to present tne methodology unique to the analysis of the reactivity anomaly events and is therefore independent of the analytical methodology presented in each of the referenced documents.
Regardless of the approval status of these documents, NRC-approved methodology will be used to develop the inpute naco9eary to perform the analyses presented in RXE-91-002.
Thus, the approval of the analytical methodologies presented in the referenced reports should not be a prerequisite to the acceptance of the methodology presented in RXE-91-002.
Attachment to TXX-92093 Pago 2 of 32
~'e 2.
Question Discuss the validation of the hot-spot calculation under the fuel msit conditions that occur in the rod ejection accident.
In this case, how is the fission gas release determined from the fraction of fuel melt?
Answer TU Electric utilizes NRC-approved material properties and heat transfer mechanisms for the hot-spot calculations.
The application of the fuel material properties in conjunction with the-use of RETRAN-02 constitutes sufficient validation of the hot-spot model under the conditions of fuel molt.
The fuel material properties are extracted from the information provided in Reference 74.
The specific fuel material properties used in the hot-spot model include the enthalpy, specific heat capacity, thermal conductivity, and melting temperature.
Included within the development of these fuel material properties are the changes that occur as a result of fuel melt and the effects of fuel exposure on the fuel melting temperature.
The conduction and convection heat transfer models within RETRAN-02 have buen approved by the NRC for use as stated in References 21 ard 22.
Tneue heat transfer models in conjunction with the fuel material properties and a conservative fuel pellet power generation profile are used to calculate the energy deposition and heat transfer for each of the ten concentric fuel regions (see Figure 3.7-1].
When the fuel temperature within a region satisfies the temperature criterion for fuel melt, the amount of energy deposited in the fuel region and the amount of energy transferred to the next fuel region are calculated based on
Attachment to TXX-92093 Page 3 of 32 the phase change characteristics specified by the input material properties.
The fission gas release from the melted fuel is used to derive a portion of the source term needed to determine the offsite radiological dose.
The specifications in Appendix B of Regalatory Guide 1.77 [ Reference 14), with the limitations specified in the CPSES-1 FSAR, are used to determine the extent and type of fission gas release from the melted fuel.
For conservatism, any fuel region that attains or exceeds the fuel molting temperature is assumed to be fully melted.
Additional conservatism is included in the offsite radiological dose calculation by assuming that 50% of the iodine contained in the melted fuel is available for release from the plant secondary.
3.
ouestion The fuel rod gap conductance and radial power distribution affect the transient moderator and Doppler feedbacks and the margin to fuel enthalpy and DNBR limits.
How will these parameters be determined to insure conservative CPSES-1 licensing analyses?
Answer The fuel rod gap conductance used for the system thermal-hydraulic analysis of the reactivity anomaly events is set in accordance with the results of various sensitivity cases.
These sensitivity cases are used to establish the direction for conservatism, i.e.,
minimum or maximum fuel rod gap conductance, for each of the specific event analyses.
Attachment to TXX-92093 Page 4 of 32 The core radial power distribution is not explicitly modelled as part of the point kinetics solution to the core power response for the reactivity anomaly event analyses.
Instead, the feedback resulting from the point kinetics solution to the problem is conservatively modelled by selecting limiting Doppler and moderator reactivity coefficients or defects for the event analyses.
The methodology described in Section 4.2 of RXE-91-002 is used to calculate the Doppler and moderator reactivity coefficients and defects.
For some events, such as the control rod ejection event, it is often desirable to more accurately model the Doppler reactivity feedback while still maintaining an overall conservative feedback response.
The core power distribution is an important contributor to the overall transient reactivity feedback due ta the non-uniform effect of the ejected control rod on the reactor core power.
A Doppler Weighting Factor (DWF) is employed in these instancoc to correct for the increased Doppler feedback associated with the spatial effects of a non-uniform fuel temperature rise.
The calculation of a DWF is performed in accordance with the rcthodology described in Section 4.8.3 of RXE-91-002.
4.
Ouestion What uncertainty allowance will be included in the temperature feedback coefficients, control rod worths, boron worth, and power distribution peaking factor input to the CPSES-1 reactivity anomaly licensing analyses?
=
i Attcchmnnt to TXX-92093 I
Paga 5 of 32
-Answer
^
The appropriate limiting values for the Moderator Temperature Coefficient (MTC) protected by the Technical Specifications and the Core Operating Limits Report are used in all reactivity anomaly analyses, except for the control rod drop event analysis.
The control rod drop event analysis uses the calculated value of MTC with an uncertainty no less than that approved in Reference 1.
The uncertainty applied to the Doppler temperature feedback and to the boron worth is 10% as approved in Reference 1.
The uncertainty applied to the calculated control rod worth is determined on a cycle-specific and an event-specific basis.
The differential control rod worth and the ejected control rod worth use an uncertainty of 15% to increase calculated control rod worth.
The trip reactivity, including the effect of a stuck control rod, is decreased by 10% from the calculated nominal value.
For tha control rod drop event analysis, the inserted control bank worth is conservatively calculated with 1) the control banks at their full power insertion limit, and 2) a power distribution corresponding to an axial offset at the upper end of-the n 'rmal operating axial offset bands.
No uncertainty is app led to the worth of the dropped control rod because the dropped' control rod worth is the independent variable used to parameterize the post-drop to pre-drop Fw ratio.
l l
The augmentation factors described in sections 4.2 and 4.8.3 of RXE-91-002 are applied to the power distribution peaking factors.
Since this analytical approach implicitly assumes the power distribution peaking factors are at the licensed l
limit for the time of maximum peak during normal operation, no additional uncertainty is requirou.
The input to the control rod drop event analysis is even more conservative in l
Attcchm:nt to TXX-92093 Pcgo 6 of 32 that it effectively assumca that the cora is at the Fw limit at each cycle exposure.
5.
Question How do the TUE reactivity event analysos of RXE-91-002 differ from the CPSES-1 FSAR analyses with respect to initial / boundary conditions and system performanco?
Answer The initial / boundary conditions used in the TU Electric reactivity anomaly event analyses are essentially the same as the CPSES-1 FSAR analyses.
The system performance of the major RCS parameters, i.e.,
core power, RCS pressure, and core average fluid temperature, ere also analogous to the CPSES-1 FSAR analyses.
Although the general trend of each event is similar to that of the corresponding CPSES-1 FSAR analysis, a direct numerical comparison of the analyses is not appropriate.
The reactivity anomaly event analyses presented in RXE-91-002 employ the analytical methodologies developed by TU Electric, while the CPSES-1 FSAR analyses utilize methodology developed by Westinghouse.
The TU Electric and Westinghouse reactivity anomaly event methodologies have many similarities.
However, a few significant methodology differences do exist, as noted below.
1.
The DNBR results presented in the CPSES-3 FSAR utilize the Westinghouse W-3R correlation while the TU Electric DNBR results utilize the TUE-1 correlation.
A one-to-one comparison of the DNBR results is therefore not meaningful.
Instead, a more meaningful comparison is made by stating that those events in which DNB is I
1
-Attachment to TXX-92093 Page 7 of 32 precluded in order to satisfy the acceptance criterion, the MDNBR remains greater than the applicable correlation limit.
For those events in which the specified correlation limit is violated, a conservative estimate of the number of failed fuel pins is used to determine the resulting offsite radiological consequences.
These offsite radiological consequences are then shown to be within the acceptance criterion specified for the event of interest.
2.
The development of N-16 protection system trip setpoints used in the TU Electric event analyses use the TUE-1 DNB correlation (References 5 and 6) in conjunction with the TU Electric N-16 setpoint methodology [ Reference 3].
The CPSES-1 FSAR event analyses use the Westinghouse W-3R correlation in conjunction with the Westinghouse N-16 setpoint methodology to derive the setpoints.
A different trip setpoint affects the event results by changing the time (and hence the system conditions) at which the reactor trip occurs and, potentially, by changing the trip function providing reactor protection.
A more meaningful comparison is achieved by stating that the results of each analysis are within the acceptance criterion specified for the event of interest.
3.
The reactivity anomaly event analyses performed by TU Electric use a pcint kinetico mdac1 to predict the reactor core power response.
Several of the event analyses (e.g., control rod ejection and single control rod withdrawal) presented in the CPSES-1 PSAR use a one-dimensional kinetics model to determine the core power response.
The development of input parameters for each kinetics model is sufficiently different that
Attechm:nt to TXX-92093 LPrgo 8 of 32 small-differences in the predicted event response may exist due to the application of specific conservatism.
6.
Question Justify the assumption that the maximum (full power) Fm -
statepoint provides the most limiting DNBR statepoint for the misaligned control rod analysis.
For example, do statepoints having higher rod worths and/or maximum excess reactivity-provide a closer approach to the DNBR limit?
i Answer Because the misaligned control' rod event is a static event, the only parameter of importance is the power distribution resulting from the misalignment.
Th's power distribution is significantly influenced-by the power peaking of the unperturbed case.
Excess reactivity at a statepoint influences the absolute value of the peak power, primarily through moderator reactivity feedback.
However, the increase in peaking as a result of the misaligned control rod is relatively_ insensitive to moderator reactivity feedback.
Hence, the effects of' excess' reactivity'are.
included in the_ determination of the reference statepoint based on the maximum full power Fm statepoint.
l The combination of a high control rod worth and a high Fa may result in a more limiting analysis, especially when assuming the misaligned control rod to be withdrawn.
In addition to the statepoint with the overall maximum full power Fa, the analysis must evaluate each-full power L
statepoint with a local maximum Fm and the statepoint with the maximum inserted control bank worth to ensure the identification of the limiting statopoint.
l
Attachment to TXX-92093 page 9 of 32 i
7.
QuestiRD Describe the screening calculations used to identify the most limiting misaligned control rod and fuel assembly replacement.
How is the uncertainty in these calculations accommodated in the TUE methodology?
Answer The most limiting misaligned control rod is identified by analyzing each control rod misalignment with a two-dimensional nodal model using SIMULATE-3 (Reference 16).
These calculations are performed for rodded and unrodded configurations allowed at operating conditions.
The pin power distributions of each of the resulting radial slices are combined using an appropriate power sharing for each slice to provide an estimate of the radial peaking factor for the misaligned case.
Each of the potentially limiting misaligned control rods is then evaluated with a three-dimensional model using SIMULATE-3, thereby alleviating the need for a " screening" uncertainty.
For the mis 1caded fuel assembly event, each misloaded fuel assembly considered is analyzed with a two-dimensional nodal model using SIMULATE-3.
The resulting assembly relative power distribution is compared to an assembly relative power distribution generated for the correctly loaded core.
The resulting assembly-wise differences are evaluated, using the criteria outlined in the response to Question 8, to determine which misloadings would be detected.
Non-detectable misloaded fuel assemblies are evaluated, using the two-dimensional model, to determine the consequences of full power operation with the misloaded assembly.
The uncertainty associated with the screening calculations is
Attachment to TXX-92093 Pago 10 of 32 accommodated by evaluating, in detail, all potential misloadings which are calculated to approach the event-specific acceptance criteria using the three-dimensional nodal model.
8.
Question What criteria are used to determine if a fuel misloading would be detected?
Describe how instrument uncertainty, power tilts and failed detectors are accounted for.
Answer The relative power distribution for the misloaded fuel assembly scenario of interect is compared to the relative power distribution for the correctly loaded core design.
This comparison determines the predicted difference for instrumented fuel assemblies.
Any scenario which results in any instrumented fuel assembly exhibiting a predicted difference in relative power greater than the acceptance criterin used to evaluate a flux map is considered to be detected.
The instrument uncertainty is small in comparison to the flux map acceptance criteria.
Therefore, the analysis doer, not consider an additional penalty for the instrument uncertainty.
Although quadrant power tilts are another means of detecting a'misloaded assembly, the analysis does not credit them.
Any design asymmetries are explicitly modeled for the full core calculations and are therefore reflected in both the calculated nominal and perturbed power distributions.
Attachment to TXX-92093 Pago 11 of 32 The analysis assumes no failed detectors because the initial flux map is obtained immediately after completion of the refuelling outage.
All necessary maintenance on the flux mapping system would have been performed at this time.
9.
Ouestion How is the effect of the fuel burnup dependence of the assembly reactivity accounted for in the selection Of the limiting fuel assembly misloading?
Are cycle depletions performed for all potential misloadings?
Answer The effects associated with fuel burnup are considered for non-detectable mioloaded fuel asscmbly scenarios in which one of the assemblies contains burnable absorbers.
Those scenarios are screened by performing a cycle depletion calculation using the two-dimensional model.
Depletions past equilibrium xenon conditions are not required for other potential misloadings.
10.
Ouestion How is a limiting power distribution determined for the misaligned control rod and misloaded fuel assembly analyses?
Are the neighboring assemblics initially operating at the DNBR limit?
ADswer The relative power of each fuel pin is calculated with the three-dimensional nodal model.
These calculated relative
Attachmsnt to TXX-92093 Paga 12 of 32 4-pin power distributions are then increased with the-augmentation factor discussed in Section 4.2 of RXE-91-002.
The limiting power distribution is the distribution resulting in the maximum relative pin powers.
Neighboring fuel assemblics are not acsumed to be initially operating at the DNBR limit.
11.
Question In the calculation of the physics parameters for the hot-zero-power control rod withdrawal analysis, in what sense are the xenon-free beginning-of-cycle and equilibrium-xenon end-of-cycle conditions bounding?
Answer The xenon-free beginning-of-cycle and equilibrium-xenon end-of-cycle conditions are not intended to be bounding.
Instead, these conditions are selected to reflect the conditions prior to startup. -The conservatism present in
-the core physics calc '. cions comes from not crediting the Doppler; reactivity feedback effects resulting from the fuel
-temperature-increase prior to the predicted time of minimum DNBR. fin addition, the differential control rod worth-and core. peaking factors calculated assuming a control rod bank overlap of 100% are significantly greater than the-corresponding values calculated using the~ nominal overlap.
O
Attcchm:nt'to TXX-92093' a: Pcg3 13Hof 32-12.
Ouestion How is the limiting rod determined for the single rod withdrawal analysis?
At what power is the single rod withdrawal event analyzed?
E_1ECI The-limiting control rod for the single control rod withdrawal event analysis is selected based on calculations similar to those used to determine the limiting control rod for the misaligned control rod event analysis (see the response to Question 8).
The basis for the selection differs in that the limiting control rod for the single control rod withdrawal event is the control rod resulting in the maximum number of fuel pins experiencing DNB.
This control rod may not correspond to the control rod resulting in the greatest Fw, as used for the misaligned control rod event analysis.
The power level used as input to the DNBR analysis for the single control rod withdrawal event is determined from an interpolation / extrapolation of the control rod bank withdrawal at power event results, Because the TU Electric control rod withdrawal event methodology uses a point kinetics solution to model the core average power response, the predicted power response resulting from the withdrawal' of a single control rod will be identical to that resulting from the withdrawal of an entire-control rod bank at an equivalent reactivity insertion rate.
The control rod withdrawal at power event is analyzed for a matrix of event scenarios that' include severcl initial pcwer levels, a wide range of reactivity insertion rates, and a variety of reactivity feedback combinations.
The single control rod withdrawal event analysis utilizes these event results to L
Attcchmont to TXX-92093 Pago 14 of 32 determine the time at which the peak core average power occurs as a function of initial power and reactivity insertion rate.
The core inlet temperature, RCS pressure, and core average power level existing at the time of peak core average power are used, in conjunction with the event-specific peaking factors, as input parameters to the DNB analysis.
13.
Question Provide additional detail and qualification for the discrete ordinates method used to determine the excore response in the dropped rod analysis.
Answer Note:
The tables and references within the text of this response that are not found within RXE-91-002 are identified by alphabetic character, and are located at the end of the response to this question.
The discrete ordinates calculations use the GIP, ANISN, and DOT 4.3 (References A, B,
and C, respectively] computer codes with the ELXSIR cross sections (Reference D].
The calculations are performed _in two steps: 1) qualification of methods by performance of the Pool Critical Assembly Problem (PCA) (References E and F] and 2) calculation of the total flux at the excore detector location for CPSES-1 Cycle 1.
~
Data describing the PCA are given in Reference E.
'These data include absolute source spectra and material and geometry. descriptions.
References E and F contain measured results for various nuclear reactions used in pressure vessel dosimetry.
Three-dimensional results are synthesized
' Attachment to TXX-92093 Pcg3 15 of 32 using DOT XY, DOT XZ, and ANISN Z calculations.
The methodology for synthesization is described in References G and H.
Tables A and B compare the results of the PCA calculations performed by TU Electric and the measured data from Reference E.
The CPSES-1 Cycle 1 excore detector response calculations
-used the same methodology as the PCA calculations.
The CPSES_ DOT calculation employed R,0 geometry to accurately model excore materials such as the pressure vessel.
DOT adjoint calculations were performed to determine a total flux response function at the excore detector location.
Determination of the total flux required transforming the X-Y pin by pin relative power distributions from SIMULATE-3
~
(Reference 16) into the CPSES DOT model using the DOTSOR code (Reference I) and folding the source function from DOTSOR with the adjoint response using the TIMEPATCH code (Reference J).
Excore detector tilte for the dropped control rod conditions were determined as the ratio of the perturbed total flux (rodded condition) to the unperturbed total flux (unrodded condition).
Calculations _were performed for 22 configurations (11 BOC and 11 EOC).
Excore detector tilts determined in'this manner were used to confirm _the algorithm described in Section 4.7.3 of RXE-91-002.- The average difference between the discrete ordinates results and the algorithm results for the 22 cases is 1.3 percent.
l
Attachment to TXX-92093 Page 16 of 32 Table A l
PCA Results Comparison of Calculated and Measured Nuclear Reaction Rates:
Np-237(n,f), Flux > 1 McV, Flux > 0.1 MeV, and DPA'.
NP-237(n,f) 2 MEASURED CALCULATED C/M TSF) 8.371E-06 7.05995E-06 0.84343 PVF 2.939E-07 2.48906E-07 0.84694 1/4 T 1.174E-07 1.13199E-07 0.96402 1/2 T 6.547E-08 6.18091E-08 0.94410 3/4 T 3.411E-08 3.33432E-08 0.97746 Flux > 1 MeV MEASURED CALCULATED C/M TSF 3.71000E-06 3.37922E-06 0.91084 PVF 1.33000E-07 1.16442E-07 0.87550 1/4 T 4.30000E-08 4.37263E-08 1.01689 1/2 T 2.07000E-08 2.04506E-08 0.98795 3/4 T 9.11000E-09 9.32052E-09 1.02311 FLUX > 0.1 MEV MEASURED CALCULATED C/M TSF 6.940E-06 5.90870E-06 0.86437 PVF 2.490E-07 2.09808E-07 0.84260 1/4 T 1.390E-07 1.31693E-u/
0.94743 1/2 T 9.350E-08 8.68525E-08 0.92890 3/4 T 5.570E-08 5.55452E-08 0.99722 DPA (BARNS)
MEASURED CALCULATED C/M PVF 1.940E-04 1.68955E-04 0.87090 1/4 T 7.510E-05 7.00332E-05 0.93253 1/2 T 4.270E-05 3.86471E-05 0.90508 3/4 T 2.260E-05 2.16791E-05 0.95925 8
DPA, displacements per atom.
2 C/M, Calculated / Measured.
3 Column 1 gives the experimental measurement locations:
TSF outer face of the thermal shield:
PVF inner face of the pressure vessel simulator; 1/4 T one fourth depth of the pressure vessel simulator; 1/2 T one half depth of the pressure vessel simulator; 3/4 T three fourths depth of the pressure vessel simulator.
~Attcchmont to TXX-92093 Pag 3 17 of a -
Table-B PCA Results Comparison of calculated and Measured Reaction Rates:
Ni-58(n,p), Al-27(n,a), In-115(n,np), and U-238(n,f)
NI-58 (n,p)
MEASURED CALCULATED C/M' 2
TSF 6.03860E-07 5.25404E-07 0.87008 PVF 2.39800E-08 2.00522E-08 0.83620 1/4 T 5.50450E-09 5.02819E-09 0.91347 1/2 T 2.18000E-09 3.96897E-09 0.90320
.3/4 T 7.74990E-10 7.65009E-10 0.98712 AL-27 (n, alpha)
MEASURED CALCULATED C/M TSF 5.40310E-09 3.98931E-09 0.73834 PVF-3.08850E-10 2.23617E-10 0.72403 1/4 T 7.06450E-11 5.67332E-11 0.80307 1/2 T 2.84000E-11 2.31291E-11 0.81441 3/4 T 1.05790E-11 9.17348E-12 0.86714 IN-115(n,np)
-MEASURED CALCULATED C/M TSF 1.013E-06 8.86838E-07 0.87542 PVF 3.629E-08 3.07148E-08 0.84642 1/4 T 1.072E-08 1.03376E-08 0.96466 1/2 T 4.971E-09 4.66913E-09 0.93933 3/4 T 2.155E-09 2.08667E-09 0.96847 U-238 (n,f)
MEASURED CALCULATED C/M 1/4 T-1.864E-08 1.63920E-08 0.87961 1/2 T 8.204E-09 7.00725E-09 0.85407 3/4 T 3.385E-09 2.95540E-09 0.87296' 8
C/M, Calculated / Measured.
3 Column 1 gives the experimental measurement 1.ocations :
TSF outer face of the thermal shield:
PVF inner face of the pressure vessel simulator; 1/4 T one fourth depth of the pressure vessel simulator; 1/2 T one half depth of the pressure vessel simulator; 3/4 T three fourths depth of the pressure vessel simulator.
L L
Attcchm3nt to TXX-92093 P ga-18 of 32 References
-A.
Lc A. Hassler Lnd N. M. Hassan, " GIP-Group Organized Cross Section Ir.put Program Manual,"
Babcock & Wilcox Report No.
NPGD-TM-456, Rev.
6, March 1985.
B.
L. A. Hassler and N.
M.
Hassan, "ANISNBW - A One dimensional Discreto Ordinates Transport Code,"
Babcock & Wilcox Report No. NPGD-TM-491,.Rev.
6, September 1987.
C.
L. A. Hassler and N. M. Hassan, " DOT 43-Two Dimensional Discrete Ordinates Transport Code (B&W Version of RSIC/ORNL Code DOT 4.3),'
Babcock & Wilcox Report No. NPD-TM-24, Rev.
2, October 1987.
D.
R.
E. Maerker, et al.,
"The ELXSIR Cross Section Library for LWR Pressure Vessel-Irradiation Studies: Part of the LEPRICON Computer Code System,"
EPRI Report No.-EPRI NP-3654, Sep*. ember 1984.
E.
W.
N. McElroy, ed.,
" LWR Pressure Vessel Surveillance Dosimetry Improvement Program: PCA Experiments and Blind Test," NUREG/CR-1A61 (HEDL-TME-80-87), July 1981.
F.
F. W.
Stallman, et al., " Reactor Calculation Benchmark PCA Blind Test Results," NUREG/CR-1872 (ORNL/NUREG/TM-428),
January, 1981.
G.
R.
E. Maerker, M.
L. Williams, and B.
L.
Broadhead,
" Accounting for Changing Source Distributions in Light Water Reactor Surveillance Dosimetry Analysis," Nuc. Sci. Eng.,
RA, 291, 1986.
l
Attachment to TXX-92093 Page 19 of 32 H.
R.
E. Maorker, et al.,
" Application of the LEPRICON Methodology to the Arkansas Nucleeir one, Unit 1 Reactor,"
EPRI Report No. EPRI NP-4469, February 1986.
I.
M.
L. Willia.ns, "DOTSOR: A Module ir. the LEPRICON Computer Code Syscem for Representing the Neutron Source distribution in LWR Cores " EPRI Research Project 1399-1, Interim Report, December 1985.
J.
R.
E. Macrker, M.
L.
- Williams, B.
L.
Brondhead, "TIMEPATCil :
A Module in the LEPRICON Computer Cc'de System for Evaluating Effects of Time-Dependent Source Distributions in PWR Surveillance Dosimetry," EPRI Research Project 1399-1, Interim Report, December 1985.
?Attcchm3nt to TXX-92093 Pcga 20 of 32 14.
Question Why isn' t-- the Doppler coefficient included as core physics input in Figure 4. 7-1 ?
Is a conservative Doppler coefficient used?
Answer Figure 4.7-1 illustrates the TU Electric control rod drop analytical methodology in terms of the parameters varied for each scenario analyzed.
As such, the variables listed under the heading of " Core Physics Parameters" coincide with the parameters used to characterize a specific control rod drop scenario.
The same-conservative Doppler coefficient is assumed for each control rod drop scenario; therefore, the Doppler coefficient is not listed among these variables.
15.
Q_uestion How is the limiting rod determined in the dropped rod analysis,-and how are multi-rod drops treated?
Answer Both single-and multiple control rod drops are considered when determining the Fat 37 and the excore detector tilt for each control rod-drop scenario.
The control rod drop event analysis' utilizes a conservative curve of excore detector tilt as a function of dropped controlcrod worth.
This curve L
bounds all excore detector tilts resulting from a control rod drop event, including those from multiple control rod I
drops.
The use of a bounding excore detector tilt function not only removes the location and exposure dependency from the calculation of_the excore detector tilt, but also l
L
Atttchment to TXX-92093 POg3 21 of 32 removes the dependence on the number of control rods dropped.
The excore detector tilt function is coupled with a model of the Nuclear Instrumentation System to produce a conservative excore detector responce during the system thermal-hydraulic analysis.
Consequently, the generation of the generic statopoints conservatively accounts for the number of control rods involved in a specific control rod drop scenario.
All unique, plausible control rod drop combinations are considered in the event analysis, including drops of one, two, three, and four control rods from each control rod group.
Each control rod drop combination is evaluated to ensure the DNBR acceptance criterion is met.
The DNBR acceptance criterion is satisfied for the event by demonstrating that Fuur is less than Fanusi for each specific event scenario.
The event scenario exhibiting the minimum margin between Fausi and Fau7 is characterized as the limitinn case.
9 16.
Question In tha control rod drop analysis, how is tho adn}r?$onal.
~
1 uncertainty due to the error introduced in the 41 1
l matrix-method, by interpolating on the input dropped rod worth, inserted bank worth and the moderator temperature coefficient, accounted for?
Answer The interpolation using the matrix of generic statepoint parameters is performed as part of a screening process to 1
Attachment to TXX-92093 page 22 of 32
+
determine the most limiting control rod drop scenario.
The physics parameters for this scenario, as identiflod by the screening process, are then used as input to a system T-H analysis to determine the exact set of corresponding statopoints.
These specific statopoints are then used as input to a core T-il analysis to determine the Fuum f or the scenario, and the corresponding margin with respect to Futn.
Any error introduced by the interpolation scheme is removed by performing the specific statopoint analysis.
Therefore, the additional uncertainty introduced by interpolation is not included as part of the analysis methodology.
17.
Question In the control rod drop analysis, the SCU methodology considers the uncertainties in tha core power., inlot temperatura and pressura to be indopondent and combinos those uncertainties statistically.
In fact, thoso calculated variablos and thoir uncertainties are couplad th!ough the noutronic/thorma: hydraulic dynamics of the control rod drop transiont and are not indopondant.
The calculation of the MDNBR uncertainty factor using the Appendix A SCU method should amploy variables yhoso uncortainties are indopondent.
The application of tho SCU method should include tho uncertainties used and their bases.
Answer The responses of the core power, temperature, and pressure are coupled through the neutronic/ thermal-hydraulic dynamics of the system.
However, the uncertainty of each variable due to steady-state fluctuations, measurement uncertainties, etc., is independent of the thermodynamic state of the
LAttCchm:nt to TXX-92093 Peg 3 23 of 32
+
'a system.
Furthermore, those uncertainties are well quantified with well-defined distributions.
Therefore, the Statistical Combination of Uncertainties (SCU) methodology as described in Appendix A of RXE-91-002, does employ uncertainties which are independent of the transient response.
The application of the SCU method in a licensing submittal will include a description of the uncertainties used and their bases.
The calculations in RXE-91-002 use the current CPSES-1 licensing basis uncertainties for the process parameters to demonstrate the use of the SCU methods.
18.
Question Since DNBR is a required prediction for the rod 0;)ection transient, why aren't the Rcs temperature and pressure selectod conservatively?
Answer The RCS temperature and pressure selected for use in the core thermal-hydraulic (i.e.,-DNB) analysis of the control rod ejection event are selected in a conservative manner.
As stated on page 4-47 of RXE-91-002, "the limiting system T-H analysis conditions for core inlet temperature and RCS pressure" are used as inputs to the DNB analysis.
This statement is not to be interpreted as a contradiction to the statements provided on page 4-40 of RXE-91-002, regarding the initial RCS fluid conditions, but rather as a clarification of the statements.
Attachment to TXX-92093 Pago 24 of 32 19.
Quentinn The treatmont of spatial offecta in reducod-dimenolon analyaos oi tho control rod ajoction accidont ia oxtremoly complex and must bo validated by co.mpariaan ta s?atial kinotics calculationa.
Damonalrato that tho TUE CPSES-1 point kinetica analysia or the rod ejection trancient is conservmtivo relativo to a detailed apatial kinotics solution.
bnrustt The treatment of spatial offects in reduced-dimension analysen of the control rod ejection accident is indeed an extremely complex incue.
The validation of the reduced-dimension control rod ejection event analynes by comparison to apatial kinetica calculations han bnen documented in References 26 and 27.
The conservatiom inherent to the reduced-dimension analysis in the direct result of the derivation and/or nelection of the input parametern.
For the control rod ejection event, the main contributora to the conservatism of the analytical results are:
1.
The calculation of the ejected rod worth; 2.
The calediation of the Doppler reactivity feedback; 3.
The calculation of the cote peaking factorn; and, 4.
The nelection of bounding valuen, an a function of core exponure, to represent other input parameters.
Each of theno parameters is developed in accordance with the methodology described in RXE-91-002.
The calculation of the ejected rod worth entaila the use of multi-dimension core physica analysen in conjunction with conservative analytical assumptions.
These assumptions
i Attcchment to TXX-92093 Pega 25 of 32 i
include the uso of adverso axial power distributions (at HFP) to increase the worth of the ejected control rod, the
-consideration of potential control rod misalignments to increase the worth of the ejected control ad, and the use of an augmentation factor of 15% to further increase the calculated ejected control rod worth.
The Doppler reactivity foodback model used in the analysis of the control rod ojnction event is separated into two distinct parts.
TO Electric utilizos multi-d...nsion coro physics analyses to computo cach portion of the Doppler reactivity foodback.
The first part of the model involves calculating the coro averago Doppler reactivity foodback.
i
~
This calculation is performed in accordanco with tho mothodology described in Section 4.2 of RXE-91-002.
The j
system T-H analysis conservatively assumes a minimum Doppler f
reactivity defect, as a function of coro exposuro, from HzP l
to HFP.
i The second portion of the Doppler reactivity foodback model i
involves calculating a Doppler Wolghting Factor (DWF).
The f
I DWP is calculated in accordance with the methodology described in Section 4.8.3 of RXE-91-002.
The conservatism inheront to the use of the :alculated DWF derives.from the I
fact that the DWF is calculated based on the ejected control rod worth prior to augmentation.
The caleplation and use of the core peaking factors is the l
most important contributor to the overall. conservatism of j
the reduced-dimension analysis.
The calculation of tho total core peaking factor, F, includes two forms of q
conservatism.
Tho first form of conservatism results from-the use.of' steady-state thermal-hydraulic foodback in the calculation instead of a-transient thornal-hydraulic foodback.
This approach derives no bonorit from the t
~,~.,_.,
Attcchment to TXX-92093
- a page 26 of 32 t
thermal-hydraulic foodback resulting from the redistribution of power during a control rod ejection event that i
subsequently reduces the calculated transient F.
Tho
[
q second form of conservatism involvos the application of a
[
l very conuorvative augmentation factor to increano tho calculated Fq value.
The method used to augment the calculated F is described in Section 4.8.3 of RXE-91-002.
q In addition to tho conservatism present in the calculation of the Fn, the application of this value to the system T-il
[
analysis and the hot-spot analysis-is also performed in a i
conservative manner.
Tho application of the calculated Fn to the syctem T-il analysis is related to the uso of the DWF.
-Becauno the DWF corrects for the increased Doppler reactivity foodback associated with the spatial offects of a non-uniform fuel temperature riso, an increano to the calculated F results in a corresponding increase to tho q
DWP.
An increaso to the DWF suissequently results in a lower predicted value for the peak power and a lower power history
- curvo, i.n.,
a lower value for the integrated full power _
i seconds (FPS).
Thoroforo, the calculated Fu prior to augmentation is used to calculate the smallost DWF for the specific scenario of interest.
The augmented F is then u
used in the hot spot analysis to determino the extent of fuel molt.
Thn application of the calculated peak Fq value to the hot-spot analysis is also conscrvative with respect to the multi-dimension analysis.
The TU Electric methodology for j
the hot-spot analysis includes the assumption that the pro-ojected peak Fu and the post-cjected peak Fq occur at exactly the same core location.
This assumption-is l
conservative because multi-dimension analysis of the redistribution of power resulting from the ejection of a control rod, finds the location of the post-ojected peak Pn
Attachmont to TXX-92093 Pago 27 of 32 l
i to bo different from the location of the pro-ojected peak F.
The assumption that the pro-ajected peak P and the n
q post-ojected peak r are situated at exactly the same coro n
location guarantoos that the hot-spot analysis is performed at the conditions of maximum initial fuel temperaturo and onorgy deposition.
4hus, the predicted fuel temperaturo and enthalpy responses will bound any other combination of pro-ojected and post-ejected total peaking factors.
The final _ conservatism employed in the uso of bounding values to characterize the many physics and thermal-hydraulic paramotors required as input to the point kinetics analysis of tho control rod ojection ovent.
This approach
[
is conservative with respect to performing a multi-dimension analysis because the detail (coro exposure, cross-sections,
- ote.) used for the multi-dimension kinetics analysis yields results.that are more representative of the ovent transient
- responso, i
The reference made to the comparisons performed by vendors (References 26 and 27) is provided to demonstrato that-similar conclusions are obtained for similar applications of reducod-dimension analysin rolovant to a multi-dimensior.
analysis.
The reference to analyses performed by other utilities (Reference 28) is provided to again demonstrate the reduction of ovent consequences'resulting from the L
application of a more dotalled multi-dimension analysis.
20.
QuestiRD How will it be insured that the Rotoronco 26 RCS overprossurization analysis for the rod ojection transis nt romains valid for futuro CPSES-1 cyclo roloads?
<!" w F T%-Ws W-
-w
-1't w-F 1"'y--freCS-r
=^m"-Vg+
y-Cp*T"*e-&+r-ve"'*M MP W%%M==owrruwq*Y-'trrg-M*-gwyT p W w + ---'ygvvyyg--+%'*-e y-Wt**pf
"(
w' W-M-"-
' s u Q -i-e W-a-g 9% wee
Attachment to TXX-92093 ~
Page 2E of 32 i
AREWar I
The intent of the generic analysin presented in Referenco 26 was to perform a control rod ojection analynis that would result in a pressure transient that would bound any anticipated control rod ojection event.
As such, the j
generic analysis utilized a very conservative set of assumptione to perform the system overproosurization analysis.
Among the moro conservative assumptions used in tho' generic analysis was the selection of an ojected control rod Worth that is more than four tiri.cs the ejected -control rod worth asnumed in the TU Electric analysis.
Bo TO Electr c will not design and licenso reload core dta23ns that could produce an ovent of such magnitudo, the j
overprossurization analysis is not included as part-of tho i
control rod ojection analytical methodology.
i 21.
Question How is the highest worth rod datormined for tha rod ojection transient?
e Answer i
The highest worth control rod for-the control rod ejection event analysis is datormined using a two-dimensional nodal calculation to estimate the ejected control rod worth.
If the results of those calculatior o are inconclusivo, i.e.,
L two or more unique control roda have a similar ejected rod worth, a throo-dimensional nodal calculation is performed to datormine which ojected control rod-results in the greatest l
ojected control rod worth.
Attcchment to TXX-92093 Pago 29 of 32 22.
OuestinD How do the section-S.4 CPSES-1 cyclo 1 calculations of tho control rod drop event compare to the y predictions?
Answer As described in Section 5.4 of RXE-91-002, the typical system response to a control rod drop event, as predicted using the TU Electric analytical methodology, is presented I
in Figures 5.4-3 through 5.4-10.
Additionally, Figures 5.4-11 and 5.4-12 depict the comparison between Fantiu and Four for postulated control rod drop scenarios initiated from BOC and EOC conditions, respectively, for CPSES-1 Cycle 1.
For each event scenario, F ny is less than Fan.uus-3 thus satisfying the DNBR acceptance criterion for the event.
The system performanco of the major RCS paramotors, e.g.,
core power, core fluid temperaturo, and RCS pressure are analogous to those of tho' analysis presented in the CPSES-1 FSAR (Reference 11).
The CPSES-1 FSAR does not provide any figures to illustrate that the DNBR acceptanco criterion is met for each control rod drop scenario, i.e.,
figures j
analogous to-RXE-91-002 Figures 5.4-11 and 5.4-12.
- Instead, L
the CPSES-1 FSAR states thnt "In all casos, the minimum DNBR i
remains above the limit value."
As a result, a direct and meaningful comparison of the Westinghouse predicted responso L
to the response predicted using the TU Electric analytical methodology is not practical.
23.
Question Discuss the variation in prompt noutron lifetimo of 1
Table 5.5-1.
What values waro used in the calculations?
- _..... -.., - ~ -.., _
.,_.n
Attcchmont to TXX-92093 Pcga 30 of 32 I
i
?
Answer The prompt neutron lifetime for the control rod ejection event analysos'prosented in Table 5.5-1 varios from 17.5 to l
29.0 microseconds.
This range is expected to bound all futuro cyclo designs for CPSES.
The values presented in
[
Tablo 5.5-1 represent the value of the prompt noutron
-lifetino used in the specified analysia.
24.
Question
'Provido tha nothodology, predictions and sensitivity studios for the control rod ejoction DNBR analyses.
Answer
[
The response to this question will be provided in a separate transmittal.
t i
25.
Question i
. In tho rod ojection accident analysis, the use of a film boiling hoat transfor correlation is consorvativo for fuel enthalpy calculations, but la nonconservative for ' heat flux prodictions in DNBR analyses.
How ia the haat transfor calculation performed in'tho DNBR analysis?
I Answer The response to this-yuestion will be provided in a separate transmittal.
L l
-_.u_..,,.
.._,_.--..u
-.. _..... _. _ _.,, _, _ _ _ -.. ~.. _.
_, _. _ _. ~.. _. - _ _ _. -....,....... _. _'
Attcchmont to TXX-92093 Pcgo 33 of 32 8-26.
Ettest1QD In the control rod ejection analysis, how is the fission gas voloaso datormined from tho fraction of fuol molt?
Answer The fission gas relcano from the molted fuel la used to dorive a portion of tho source term nooded to dotormine the offsite radiological doso.
The specifications in Appendix B of Regulatory Guido 1.77 [Roforence 14), with'tho limitations specified in the CPSES-1 FSAR, aro used to l
determine the extent and type of fission gas release from the molted fuel.
For conservatism, any fuel region that i
attains or excoods the fuel molting temperature is assumed to be fully molted.
Additional conservatism is included in the offsito radiological doso calculation by assuming that 50% of the lodino contained in the molted fuel is available for release from the plant secondary.
- m,...
m 27.
Question ~ M
~1 F
In previous analysos, the middlo-of-cycle statopaint has been found to be limiting in the ovaluation of the control rod drop ovent.
How will it bo insur3d that a middio-of-cyclo statepoint is not limiting for futuro CPSES-1 cycle I
reloada?
Anfmer-For cyclo exposures at which core physics paramotors are not explicitly generated, tha core physics paramotors are estimated by assuming a linear variation with cycle exposure between the explicitly defined conditions.
If a cyclo l
Attachmont to TXX-92093 e.
page.32 of 32 exposure other than those explicitly defined in dotermined to be more limiting, explicit calculations of the core physics paramotors are performed for that coro exposuro.
The core physics paramotors at this additional exposure and the values from the original exposure are then upod to estimato the core physica paramotors at the remaining cycle exposures by again assuming a linear variation with respect to cycle exposure.
The process is repoated, na necessary, to confirm that tho event-specific acceptanco critoria are mot for all cycle exposuros.
i 28.
Question i
a How will conservativo values of tho delayod noutron fraction and prompt noutron lifetimo bo datormined for each of tho i
AnaERE j
The delayed neutron fraction and prompt neutron lifetime used for the system thormal-hydraulic analysis of the reactivity anomaly ovoats are set in accordance with the results of various sensitivity studios.
Those sensitivity i
studios are used to establish the conservativo direction, i.e.,
minimum or maximum parameter values, for each of tho specific event analysos, f
t L
L
-