ML20126H549
| ML20126H549 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 12/14/1992 |
| From: | Carew J BROOKHAVEN NATIONAL LABORATORY |
| To: | Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20126H551 | List: |
| References | |
| CON-FIN-A-3868 NUDOCS 9301050250 | |
| Download: ML20126H549 (19) | |
Text
.
ATTACHMENT TECHNICAL EVALUATION-REPORT TECHNICAL EVALUATION OF TOPICAL REPORT RXE-91-002 (REACTIVITY ANOMALY EVENTS METHODOLOGY)
FOR TEXAS UTILITIES ELECTRIC J.F. Carew December 14, 1992 Prepared for the Office of Nuclear Regulatory Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555 NRC FIN A-3868, Task # 34 Reactor Analysis Group Advanced Technologies Division Brookhaven National Laboratory Upton, Long Island, New York 11973
_ _ , - s 1
-;KA
.~
i g
1,0 INTRODUCTION By letter dated May 31, 1991 (Reference-1) TV Electric (TVE) has submitted the Topical Report RXE-91-002, " Reactivity Anomaly Events Methodology." The proposed RXE-91-002 methodology is intended for the. analysis of the FSAR Chapter-15 reactivity anomaly events for the Comanche Peak Steam Electric System (CPSES). The methodology makes use of previously developed TUE models and results from the core reload design (References 2-4), core thermal--
hydraulic analysis (References 5-7), and system thermal-hydraulic analysis (References 8-9).
The topical report provides a description of the TUE analysis philosophy, and-event classification and acceptance criteria. The calculational tools and models used to (1) perform the calculation of the core response to the accident reactivity, (2) determine the neutronic and thermal-hydraulic input to the transient calculation and (3) evaluate the consequences of the event, are described in detail. The TUE methods, including the assumptions concerning the system performance and initial / boundary conditions made in carrying out the event analysis are also given.
The review focused on the applicability of the codes and models employed in the calculations, the validity and conservatism of the assumptions made in modeling the events, and the overall completeness of the event analyses. - The TUE methodology- and applications are summarized. in the following Section-2, and the technical evaluation of the important issues raised during the review is presented in Section-3. The technical position is given in Section-4.
. . .- - - . , ~ . . - , ., . . - . - -,- ,
y<
.*s 2:
2.0- SIM4ARY OF THE TOPICAL REEQRI The TUE reactivity anomaly events methodology consists of?(1) the' ,
neutronic/ thermal-hydraulic' models used to calculate the nre transientJ response and (2) the event-specific assumptions and analyses made to conservatively evaluate the consequences of the-events. 2The details' ofIthese aspects of the TUE methodology are sumarized in the folloJ 3 2.1 Rg. activity Event Analysis Models _
g 2.1.1 Core Neutronics Model Steady-state three-dimensional core neutronics calculations are performed with SIMULATE-3.(Reference-10). SIMULATE-3 is a two-group nodal-. diffusion program that includes a local power reconstruction capability allowing.the determination of the intra-nodal power distribution. -The nodal neutronics. .
input for SIMULATE-3 is determined with CAS 40-3 (Reference-ll) as: a function
~'
__of fuel burnup, control, and moderator and fuel temperatures. CASMO-3 calculates the fuel assembly neutronics characteristics in multigroup?
transport theory, using a det' ailed rod-wise-fuel assembly geometry model. l SIMULATE-3 is used to calculate the core-power distribution, feedback coefficients, rod / bank worths, solublo boron worth and:the-core. kinetics
- parameters. A one-dimensional one-group: diffusion theory core _model. is used to generate generic axial power _ shapes ~ for events which do not require an event-specific axial power distribution.
3 2.1.2 Core Thermal-Hydraulic Model The CPSES core thermal-hydraulic model employs VIPRE-01 (Reference-12) to calculate the thermal-hydraulic state and the core MDNBR. The reactor core is modeled as a set of one-dimensional channels that are coupled by lateral cross flow and turbulent mixing. The VIPRE-01 model includes a detailed fuel rod model which allows the determination of the fuel pellet temperature distribution and fuel enthalpy.
2.1.3 System Thermal-Hydraulic Model The CPSES system thermal-hydraulic model uses RETRAN-02 (Reference-13) to calculate the core and system response to the reactivity anomaly events. The RETRAN-02 core power response is calculated with point kinetics using moderator and fuel temperature feedback coefficients and reactivity forcing functions to describe the scram and control rod motion. The calculation includes six delayed groups and the 1979 ANS decay heat standard.
The RETRAN-02 trip control logic is used to model the trip functions that occur during the reactivity events. These include the reactor trips on high neutron flux, high and low pressurizer pressure, high pressurizer water level and the overpower and over temperature AT trips. The CPSES model includes both the pressurizer pressure control system, which limits RCS pressure changes during the transient, and the motion of the control rod drives to maintain the preprogramed average temperature.
4 2.1.4 Hot Soot Model In order to assist in the calculation of DNBR a hot spot model is also included in RETRAN-02. The hot spot model represents an axial segment of the highest powered rod in the core. The model includes explicit regions for the fuel pellet, gap and clad. The fuel pellet radial power distribution is input, and the Zircaloy-water reaction and fuel pin heat transfer coefficient !!
are calculated.
2.1.5 Boron Dilution Model The RETRAN-02 boron dilution model is used to calculate the boron concentration in the RCS during the boron dilution event. The RETRAN-02 general transport model is used to determine the system changes in boron concentration. As part of the modeling of the boron dilution mitigation system (BDMS), the RETRAN-02 control system monitors the neutron flux during
~
the event and switches the charging pump to the reactor water storage tank when the neutron flux doubles is less than ten minutes.
2.2 Reactivity Event Analyses 2.2.1 Ggneric Core Physics Parameters Kinetics Parameters The moderator and Doppler reactivity feedback coefficients are l
l
5 calculated with SIMULATE-3. The noderator coefficient is calculated by varying the inlet temperature, and is determined as a function of boron concentration, moderator temperature, fuel exposure and rod insertion. The Doppler coefficient is calculated by assuming a spatially uniform increase in the fuel temperature. The nodal delayed-neutron parameters are calculated with CASMO-3 and spatially weighted over the core volume using the global flux shape calculated with SIMULATE-3.
Core Power Distribution The radial pin-wise peaking factor F,3, total three-dimensional heat flux peaking factor F , and the core axial power shape are calculated with SIMULATE-3. An augmentation factor defined as the ratio of the full power design limit peaking to the maximum calculated power peaking is applied to F an -
when the radial peaking exceeds the nominal power peaking. Conservative event-specific augmentation factors are applied to the F, three-dimensional peaking factor.
Control / Scram Rod Worth The control and scram reactivities are determined with SIMULATE-3. The scram reactivity is calculated as the difference between the core multiplication with the rods at the rod insertion limits and with all rods inserted and the highest worth rod stuck. In order to minimize the scram worth a top skewed axial power shape and a conservative scram velocity is used. In addition, a 10% calculational uncertainty allowance is applied to the scram worth and a
4
. 6 1
15% uncertainty allowance is applied to the bank worth. '
2.2.2 Misalianed Control Rod and Misloaded Fuel Assembly i Both the misaligned control rod (MCR) and misloaded fuel assembly (MFA) events are quasi-static events involving a potential approach to DNBR limits. These !
events are calculated with the full core SIMULATE-3 model. The MCR is analyzed.
at full power and at the time in the cycle where the maximum design F a occurs. The hot channel peaking factor, including augmentation factor, is-calculated for the limiting misalignment and then used to determine the MDNBR.
Simplified screening calculations are used to identify the limiting fuel assembly misloading. The core surveillance instrumentation response is evaluated to determine the misloadings which would not be detected. The MDNBR is then calculated for these misloadings at the limiting system conditions for full power operation.
2.2.3 Boron Dilution Analysis The boron dilution event is a quasi-static transient involving a potential-approach to the core DNBR limit. The increase in core multiplication resulting from an assumed conservative boron dilution rate is calculated with the CPSES SIMULATE-3 model. A conservative dilution rate is determined by maximizing the critical boron and initial boron concentrations, and the dilution rate. The analysis must demonstrate that there is sufficient time to terminate the event prior to the loss of shutdown margin (SDM) and to minimize
.- r v
i 4
-p . 4
- this' time"a minimum RCS volume and SDM, 3nd a: maximum' purge volume a're-assumed.
-)
2.2.4 Control Rod Withdrawal Analysis The control bank withdrawal event is analyzed at hot-zero-power (HZP) Land"at- y power conditions. The HZP analysis assumes two sequential maximum worthT H control banks in 100% overlap, while the at-power analysis assumes maximum' ,
- worth sequential banks in normal overlap. The. single-rod withdrawal is.
- analyzed at the RIL limit. The RCS pressure response and ADNBR are maximized by maximizing the initial RCS temperature and reactivity insertion-rate and -
minimizinig the initial RCS flow. The RETRAN-02 model is used to determine the--
system response, which is used to define the boundary conditions for thei detailed DNBR analysis.
2.2.5 Control Rod Droo Analysis Y
The control rod drop event involves _an' initial reduction in core 1poweridue to J the dropped rod, and_ a:potentialf oyershoot of the initial. power and approach - .
- to DNBR limits as the control- banks are ' automatically withdrawn. The system:-
response to this event'is calculated with the CPSES'RETRAN-02 inodel. The J accident ADNBR is makimized by using a maximum RCS temperature and minimum'RCS:
pressure, flow and Doppler feedback. The-SIMULATE-3 model-is used to-- ,
determine the limiting control rod worths,- feedback reactivities.and-hot channel peaking factors.--The dropped rod reactivity insertion is assumeo to u
be' instantaneous,:and top peaked' power distributions are used to maximize the ;
( -
![
l e .
o
d g.
withdrawn bank worth. The VIPRE-01 core thermal-hydraulic model, together with a statistical combination of uncertainties (SCU) method, is used to calculate the margin to DNB.
2.2.6 Control Rod E.iection Analysis The core response during the control rod ejection (CRE) event is calculated with the RETRAN-02 system model. In order to insure an analysis that bounds l the expected operating conditions, a maximum ejected rod worth and a minimum fuel temperature feedback and delayed neutron fraction are employed. The-analysis is performed at beginning and end-of-cycle, and at full power and hot-zero-power conditions. The RETRAN-02 physics parameter input, including reactivity feedback, scram and ejected rod worth, and power distribution peaking factors, is determined with the CPSES SIMULATE-3 core model. A 15%
j uncertainty allowancc is included in the. ejected red worth. The CRE fuel hot-spot analysis assumes a maximum initial fuel temperature and film boiling heat transfer (shortly) after the rod is ejected. The CPSES VIPRE-01 model is used l
l- to determine the F,3 required for DNB, which is then used to perform the full-core fuel pin census. A separate CRE fuel pin census is performed to determine the fraction of fuel melt.
t________-_-__-____-_____-____
9 3.0
SUMMARY
OF THE TECHNICAL EVALUATION The reactivity anomaly events methodology described in the TUE RXE-91-002 topical report includes the methods for determining input and for performing the analyses and examples of typical CPSES reactivity event evaluations.
Several important technical issues were raised during the review which required additional information and clarification from TUE. This information was requested in Reference-14 and was provided in the TUE response included in References-15 and 16. The important technical issues raised during the review of RXE-91-002 are discussed in the following.
3.1 Related TUE Core Desian Methods The methodology presented in the reactivity events Topical Report RXE-91-002 expands on previously developed TUE core design methods. These include the reload core design analyses (References 2-4), the VIPRE-01 core thermal-hydraulic analysis methods (Reference-5), the TUE-1 DNB correlation (References 6-7), and the RETRAN-02 system transient analysis methods (Reference-8). These methodologies are used to provide input to the reactivity event analyses (e.g., control rod / bank worths and reactivity coefficients), are integral components of the methodology itself (e.g., the VIPRE-01 core and RETRAN-02 system models), or are used to determine the consequences of the reactivity events (e.g., the number of rods in DN8 or experiencing fuel melt). Since these supporting methodologies can have a significant impact on the analyses of the reactivity events, the application of the RXE-91-002 reactivity events methodology must be carried out with
~
10
)
approved versions of these TVE core design methods.
3,2 Static Analvigi The consequences of the reactivity anomaly events which result from local reactivity perturbations, such as the misloaded fuel assembly, control rod withdrawal and the control rod ejection events, are sensitive to the core location of the reactivity perturbation. Consequently, the determination of the worst-case or bounding events requires a preliminary static evaluation of a large number of error rods, misloaded fuel assemblies and core statepoints.
The identification of the worst-case misloaded fuel assembly is especially difficult, since the misloading affects the core exposure distribution and a core depletion is required for each fuel assembly mitloading. In the TVE methodology, the limiting error / accident rod and misloaded fuel assembly are identified by a series of two-dimensional SIMULATE-3 screening calculations.
In the response to the RAI (Reference-15), TVE has indicated that while the event screening is performed in two dimensions the effect of these local perturbations is evaluated in detail using a three-dimensional SIMULATE-3 model.
In the rod misalignment and misloaded fuel assembly static analyses TVE incorporates additional conservatism into the DNBR analysis by using a hot channel augmentation factor which increases the hot channel power to the full-power design limit (Response-10, Reference-15).
Based on the definitions of the methods and the conservatisms included, it is
\
l
, 4 l <
p.- -
~
concluded that the'TUE methods for performing thelstatic analyses;are :
acceptable. - -
3.3 Control Bank Withdrawal Analysis The control-bank withdrawal from hot-zero-power typically involves the largest'-
reactivity insertions and, depending on the reactivity insertion rate,-is -
provided DNBR protection by either the OTN-16:or high flux trip. -TUE-has.
included additional conservatism in the HZP bank withdrawal analysis by (1).
neglecting the Doppler feedback resulting from the power increase, (2)'
assuming the two highest worth banks are withdrawn, (3) taking the sequential-
~
banks to be in 100% overlap and (4) increasing the calculated Fan _:by a-conservative peaking augmentation faccor (Reference-15).
Based on the responses of Reference-15 and.the' identified conservatisms we' find the control bank withdrawal analysis to be acceptable.
il 3.4 Control Rod Droofaalysis !
l
- In the control rod drop analysis, the extent' of the power overshoot and approach to DNBR limits is determined, in part, by the initial reduction in - i power sensed by the excore detectors. The. maximum. power overshoot results from ti.e largest sensed reduction in power. In order to provide an' accurate.
p L determination of the power reduction measured by the excore detectors, TUE. has .
performed radial (r,6) discrete ordinates transport calculations of the excore n
response. The calculational methodology employed in this analysis was b
. .. - ~ . . . - .
j v -
y
.g7 I
~l 12-
- extensively benchmarked against the pool critical- assembly benchmark" experiment (Response-13,t Reference-15). A conservative excore response is- .
determined by assuming a maximum radial power tilt and al worst-case !
instrumentation response (Response-15, Reference-15).- j The uncertainty allowance in the control rod drop event is determined usingla statistical ' combination of uncertainties approach. - In Response-17 (Reference-
- 15) TUE indicates that the input uncertainty values for the SCU employed in RXE-91-002 are the CPSES-1 licensing basis uncertainties, and that future licensing submittals-will include the uncertainties assumed and their bases.
It is concluded that this approach provides an accurate and acceptable method-for the analysis of the control rod drop event.-
3.5 Control Rod E.iection Analysis The acceptance criteria for the control rod ejection event inc?ude specific-limits on the transient fuel enthalpy, offsite doses and reactor coolant-system pressure._
In order to demonstrate that the consequences'of the~CRE event are within the-specified fuel enthalpy and dose limits,.--TUE calculates -
the' fuel melt fraction, number of rods experiencing DNB and the offsite doses.-
l l- TUE does not calculate the CRE overpressurization resulting from the RCS--
I -temperature increase and potential coolant. voiding, but intends to use the CPSES FSAR overpressurization analysis to demonstrate.conformance with the RCS.
- pressure limits. While the CPSES FSAR overpressurization analysis for the CRE event is believed to be generally bounding,-the conservatism of this analysis i- -
l should be demonstrated for each CPSES reload core design.
I L
4 w n w-w s - n- m c. ~
e
13 The topical report includes detailed sensitivity analyses in support of the CRE calculations. These sensitivity analyses provide the variation in peak power, transient integral power, fuel enthalpy and fuel melt to expected variations in calculational input. In Response-24 (Reference-15) the corresponding sensitivity of the number of rods in DNB is presented, and indicates a relatively weak sensitivity for the hct-full-power analysis and larger DNB sensitivities for the hot-zero-power calculation. These sensitivities are consistent with the CRE transient dynamics.
The rod ejection analysis is performed using the RETRAN-02 point kinetics model. However, the REA results in strong spatial effects that are not explicitly included in the RETRAN-02 model. Both the local peaking and Doppler weighting factor undergo large increases, as the rod is ejected and the power distribution redistributes, which have a significant effect on the REA transient dynamics. In the response of Reference-16, TUE has compared the RETRAN-02 REA analysis to the one-dimensional Westinghouse (W) analysis included in the CPSES FSAR. These comparisons indicate that the TUE point kinetics model predictions of peak core power, peak fuel centerline temperature, peak average fuel temperature, and peak fuel _enthalpy are conservative relative to the W predictions. This conservatism is due, in part, to the conservative treatment of the scra:a reactivity following trip.
It is therefore concluded that the TUE rod ejection analysis model is acceptable for reload cores similar in design to CPSES Cycle-1, provided the conservatism in the treatment of the scram reactivity is maintained. The TUE point kinetics rod ejection analysis model will require requalification-if
14 l applied tu reload core designs which differ significantly from the CPSES Cycle-1 benchmark, or if the conservatism in the scram reactivity is relaxed..
3.6 Reactivity Methodoloov Conservatism and Uncertainty Allowinqg The consequences of the reactivity anomaly events are determined by the mismatch between the inserted and feedback reactivities. TUE intends to calculate the input reactivity parameters and coefficients using NRC approved methods. However, the calculated reactivity inputs have inhere...
uncertainties which can have a substantial effect on the results of the event analyses. To account for these uncertainties, TUE adjusts the control rod / bank worths, scram delay, feedback coefficients, boron worth and peaking factors in the direction shown to be conservative by event-specific 1
sensitivity analyses. In Response-4, TUE provides the uncertainty allowance which is to be applied to the important reactivity parameters. The indicated values have been approved previously, are technical specification limits, or are reasonable estimates of the expected uncertainty for the TUE methodology.
In Pvssonse-28 (Reference-15), TUE has indicated that conservative values for the delayed neutron fraction and prompt neutron lifetime transient parameters will also be determined on an event-specific basis.
This method provides an acceptable treatment of the effect of the reactivity input uncertainties os, the reactivity anomaly events.
i
. . . . , . , . . . - - i
n l
E' 15 4.0 TECHNICAL p0SITION The TUE reactivity anomaly events methodology-RXE-91-002, including the calculational models, event analysis methodology and typical CPSES reactivity event evaluations, has been reviewed in detail. Based on this review, it is concluded that the methofs described in RXE-91-002 are acceptable for performing CPSES reload licensing evaluations with the following limitations.
- 1) Related TUE Corg_Desian Methods The TUE core design methods of References 2-8 are used to provide input and perform calculations as cart of the proposed RXE-91-002 reactivity events methodology. Since thase supporting methodologies can have a significant impact on the consequences of the reactivity events, the application of the RXE-91-002 methods must be carried oJt with approved versions of these related methods (Section-3.1) 1
- 2) Control Rod E.iection Overoressurization Analysis A generic overpressurization analysis for the control rod ejection event is l provided in the CPSES FSAR. While this analysis is believed to be generally l
l bourding, the conservatism of the FSAR overpressurization analysis should be demonstrated for each CPSES reload core design (Section-3.5).
U f -;<
i-16
- 3) Control Rod Eiection Point Kinetics Analysis The TUE point kinetics rod ejection analysis model will: require requalification if applied to reload core designs which differ significantly from the CPSES Cycle-1 benchmark, or if the conservatism in the scram reactivity is relaxed (Section-3.5).
- 4) Statistical Combination of Uncertainties The uncertainty values and bases used in the statistical combination of uncertainties method must be provided in the cycle-specific application of the RXE-91-002 methodology (Section-3.4).
{ .
o 17 Referencn I
- 1. " Comanche Peak Steam Electric Station, Docket Nos 50-445 and 50-446, Reload Analysis Program, RXE-91-002, Reactivity Anomaly Events Methodology," Letter, Cahill, W.J. (TVE) to U.S. NRC, dated May 31, 1991.
- 2. Edwards, D.J., Kostyniak, L.E., Monger, F.A., Rubin, R.M., and Willingham, C.E., " Steady State Reactor Physics Methodology," RXE 003-P, TU Electric, July 1989.
- 3. Edwards, D. J., " Control Rod Worth Analysis," RXE-90-005, TU Electric, December 1990.
- 4. Bosma, J.T. and Grace, M.A., " Power Distribution Control Analysis and Overt:mperature N-16 and Overpower N-16 Trip Setpoint Methodology," RXE-90-006-P, TU Electric, February 1991.
- 5. Sung, Y. X, and Giap, H. B., "VIPRE-01 Core Therral-hydraulic Analysis Methods for Comanche Peak Steam Electric .9ation Licensing Applications," RXE-89-002, June 1989.
- 6. Giap, G. B. and Sung, Y. X., "TUE-1 Departure from Nucleate Boiling-Correlat son," RXE-88-102-P, January 1989.
- 7. Giap H. B. and Hiltbrand, D. W., " TUE-1 DNB- Correlation, Supplement 1," RXE-88-102-P, Sup. 1, December 1990.
- 8. Lo, S. S., Devore, C. V., and Boatwright, W. J., " Transient Analysis Methods for Comanche Peak Steam Electric Station Licensing Applications," RXE-91-001, February 1991.
- 9. Boatwright, W. J., Maier, S. M., and to, S.S., " Design Basis Analysis of a Postulated Steam Generator Tube Rupture Event for Comanche Peak Steam Electric Station, Unit 1," RXE-88-101-P, MARCH 1988,
- 10. Umbarger, J.A. and DiGiovine, A.S., "$IMULATE-3: Advanced Three-Dimeastonal Two-Group Reactor Analysis Code User's Manual,"
Studsvik/SOA - 89/03, Studsvik of America, November 1989.-
- 11. Edenius, M., Ahlin, A., and Forssen, B., CASM0-3: A Fuel Assembly Burnup Program User's Manual," Studsvik/NFA - 88/48, Studsvik of America, September 1988.
- 12. Stewart, C. W., Cuta J. M., Montgomery, S. D., Kelly, J. M., Basehore, K. L., George, T.L., and Rowe, D.S., "VIPRE-01: A Thermal Hydraulic Code for Reactor Cores." NP-25411-CCM, Revision 2, Electric Power Research Institute, July 1985.
q t
. . i-
.4-18
- 13. McFaddes4 J.H.,.Peterson, C.E., Paulsen, M.P. and Gose, G.C., "RETRAN-02, A Program for Transient Yneraal-Hydraulic Analysis of Complex. Fluid -
Flow Systems," NP-1850-CCM-A, Revision 4 Electric Power:Research Institute, November 1988.
- 14. " Request for Additional Information on RXE-91-002 Reactivity Anomaly.
Events Methodology," Letter,- T. Bergman (NRC) to W.J. Cahill, Jr.~
(TUE), dated January 14, 1992.
15.
Comanche Peak Steam ' Electric Station, Request for Additional ~
Information on RXE-91-002 Reactivity Anomaly Event: Methodology,"
Letters,W.J.Cahill,Jr.(TUE)toU.S.NuclearRegulatoryCommission, .c dated February 14, 1992 and March 31, 1992.
~
- 16. " Comanche Peak -Steam Electric Station, Request 'for Additional Information on RXE-91-002 Reactivity Anomaly Events Methodology,"-
Letter, W.J. Cahill, Jr. (TUE) to U.S. Nuclear Regulatory Commission, ,
dated October 12, 1992.
[
I i
+
Af:K 1 J+y e ,
t I
l 4
l^
L l
I
- . . . _ - , ,