ML20235D704
| ML20235D704 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 09/22/1987 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20235D694 | List: |
| References | |
| NUDOCS 8709250317 | |
| Download: ML20235D704 (6) | |
Text
__
\. p uc%
y g kg UNITED STATES NUCLEAR REGULATORY COMMISSION
% j WASHINGTON, D. C. 20555
\...*/ SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO GPU NUCLEAR COPR0 RATION TOPICAL REPORT TR-021, REVISION 0
" METHODS FOR THE ANALYSIS OF BOILING WATER REACTORS STEADY STATE PHYSICS" GPU NUCLEAR CORPORATION OYSTER CREEK NUCLEAR GENERATING STATION DOCKET NO. 50-219
1.0 INTRODUCTION
By letter dated March 25, 1986 (Ref. 1), the GPU Nuclear Corporation (GPUN) submitted for review TR-021, Revision 0, " Methods for the Analysis of Boiling Water Reactors Steady State Physics." The information in this report was supplemented by information submitted with References 8 and 9 in response to requests for additional information from the NRC staff and consultants. The review by the staff of this report and supplemental infomation was perfomed with the assistance of consultants from Brookhaven National Laboratory (BNL).
As indicated in Reference 1, it is the intent of GPUN to conduct in-house analyses for core related changes to the Oyster Creek Nuclear Generating Station (0yster Creek) Technical Specifications during Cycle 11, and perform relcad core safety analyses for Cycle 12. This report is the second of four submitted by GPUN. A report on lattice physics (TR-020) has been reviewed and approved and two reports related to the analysis of transients (TR-033 and
. 040) tre being reviewed. This report (TR-0211 describes the three-dimensional BWR steady)
(neutronic state coupled EPRI-NODE-B and r.eutronic/themal-hydraulic modeling (thermal-hydraulic) EPRI-THERM-B using codes, andthe is referred to as the N0DE-B code. The report also provides verification of the accuracy of the calculations with NODE-B by comparisons with measured data.
Both EPRI-NODE-B and EPRI-THERM-B are part of the Advanced Recycle Methodology )
Program (ARMP) code system (Ref, 4). The NODE-B three-dimensional core simulator code has been developed with the EPRI Power Shape Monitoring System (PSMS) (Ref, 5), an on-line hybrid system which monitors core performance and i power distribution. T1e integrated NODE-B/ THERM-B code system of the on-line l PSMS has been converted for use by GPUN on the IBM computer for off-line analysis. The modeling and verification of the N0DE-B/ THERM-B integrated system, used off-line, is the subject of the present technical evaluation.
2.0 DESCRIPTION
OF THE METHODOLOGY l The two ARMP codes, EPRI-NODE-B and EPRI-THERM-B, have been integrated into a single code, N0DE-B, for the PSMS application. The integrated code is a l coupled three-dimensional neutronic and themal hydraulic model in which a '
complete calculation consists of a converged set of iterations between neutron source and moderator voids.
1 2.1 EPRI-NODE-B h$
P K p
4 i .o.
The EPRI-NODE-B code is a successor to the FLARE code (Ref. 61 A modified one-group theory model is used in this code. The key input parameters in EPRI-N0gE-Baretheneutronmultiplication,k-infinity,andthemigration area, M . These parameters are derived from detailed energy and space-dependent calculations for each fuel assembly type and are entered in the nodal calculation as a function of coolant voids and exposure, including the effects of control, coolant temperature, Doppler and xenon. The fuel assemblies are coupled together in EPRI-N0DE-B using a tiansport kernel which is a function of the migration area and the nodal mesh spacing. The l transport kernel plays an important role in the nodal calculations since it, along with the local multiplication and leakage factors, is used by the code in the calculation of the three-dimensional power distribution. The code calculates the transport kernel in each node in the horizontal and vertical directions using input constants which are selected such that the results of the basic model calculations are nomalized to a more accurate calculation such as PDQ or to measured data.
2.2 EPRI-THERM-B~
This code calculates the themal hydraulic parameters of the core including flow distribution, subcooling, void and quality distributions based on total core power, recirculation flow, power distribution, and feedwater flow and temperature. Since the coolant flow distribution through the core is influenced by the void content and the power level, an iterative calculation is required to determine the power and flow distribution.
i The flow distribution is obtained by equalizing the pressure drop across each channel. This calculation starts with an initial guess for the coolant velocity in each channel and the pumphead requirements, and proceeds iteratively until the coolant velocity converges within a specified tolerance. The process is repeated for each channel. When a distribution is obtained for all of the channels, all individual channel flows are summed and compared to the total core flow. The calculation is complete when the summed flow is within a specified tolerance of the total core flow.
The subcooling in the EPRI-N3DE-B code is calculated by performing a heat balance in the downcomer and lower plenum regions of the vessel. The single-phase loss coefficients are input to EPRI-NODE-B. These coefficients are corrected during the calculation for the local quality and void condition within the channel. The relative moderator density, a key variable in the representation of the nuclear properties of the core, is detemined by calculating the nodal quality from the power and channel flow rates. The Zolotar-Lellouche (Ref. 7) void-quality model is employed in the thermal hydraulic code.
3.0 EVALUATION The evaluation of this report is based on the review of the methods underlying the NODE-B code and the verification of those methods with measured data. The material for this review includes both the rcport (TR-021) and the responses to questions (Ref. 8 and 9).
3.1 Modelina
L l
l ;
i:
. Neutronic Fodeling Mater' ial constants including the neutron multiplication and mi gration area are derive.d from multigroup fuel assembly calculations e and introduc as functions controlled of exposure and moderator voids for both uncontrol assemblies. nto NODP-B each node is based on a square root of fuel temperatur appropriate power and moderator density corrections.
These representations are known to be adequate for the analyses NCDE-B. intended to be carrie Nodal exposure effects are determined by calculating fractio changes from exposure and void dependent nal reactivities reactivity input Nodal start exposure of each is updated with each exposure step using the time step. .
the end of each time step. Exposure-weighted voids in each node are ccxiputed Control histor voids can be included in the calculation. y effects on the exposure-weighted inacceptable.
is NODE-B is sufficient for core follow This exposure reactivityanalyses and reload modeling and th erefore it the moderator density and, togetherr with functionthe and power, is of noda used in the evaluation of the xenon number density.
effect is then calculated in each node. The xenon reactivity this treatment is acceptable. effects on nodal reactivity and powe ,
Thermal Hydraulic Modeling The integration of EPRI-THERM-B with EPRI-NODE-B into a single e, N00E-B, eliminates the possibility of errors in transferring modules during the neutronic/themal-hydraulic iteration process datacodebetween t Starting with an initial guess for thennel equations is obtained iteratively. coolant aulic and an vel requirement within a specified tolerance.is evaried pumphead until it yie obtained fromtothe converged The individual channel flows flow is compared the total core flow. coolant velocities are sumed and the k specified range the problem is converged. If the two flows lie within a ng f
input and the entire iterative procedure wisin repeated.to the r The fuel assembly pressure drop is obtained as a quare function of the s liquid coolant velocity, factor and void conditions. the boiling and non-boiling lengths
, the friction of the nodal steam quality, volume which fraction. is derived from the poweresand and the channel flow neutronic/ thermal hydraulic iteration to establish constants in EPRT-N0DE-B. uclear the values The methods emp'.oyed in the thermal-hydraulic cceptable calculationsfor are a representing the steady state behavior of the Oyster Creek core .
I
A-Input Model The input model in NODE-B consists of neutronic and thermal hydraulic data.
Basic core and fuel design data, power level, control rod position nuclear constants, the input. core flow and thermal-hydraulic characteristics are spec,ified in effects are input for each fuel type. Constants needed to evaluate Doppler, xe Creek core consists of an array of cubic nodes; one node for the horizontal plane and 24 axial nodes in the axial direction. The GPUN input model is consistent with the calculational features of N0DE-B and is acceptable.
An important segment of the NODE-B input model is the data used for the {
normalization of the results to measured data.
and bottom albedos, reflector constants and partial fuel factors allows theApprop user to minimize the deviations between measured and calculated data and improve the quality of the input model. These input normalization data sets are constant throughout the analysis of the current and future cycles.
3.2 Verification of Methodology The methodology employed iri NODE-B has been verified by comparing results of calculations Creek. with measured data obtained during the operation of Oyster Both verification process. cold zero power and hot operating conditions were included in the In addition, the performance of the code was verified against measured data from Hatch 1 Cycle 1 operation includ Cycle 1. -
_ Cold Reactivity Data from Oyster Creek startup tests at the beginning of Cycles 8, 9 and 10 were used in the verification of the NODE-B cold model. The cold critical tests conducted during the startups were all local criticals. A total of 13 cold10.
and critical experiments were conducted during the startups of Cycles 8, 9 both a positive and negative period.For each local critical configuration, criticalit Calculations with the N0DE-B code yielded deviation of 0.3%.
a combined average critical k-effective based on the th shutdown margins within about 0.2% with a standard deviatio The cold model of N0DE-B is found acceptable for application to cold critical Cycles 8, 9 and 10 experiments of the Oyster Creek cycles which are similar in fuel loading t Hot Reactivity l
_E.
Core follow calculations were performed for Oyster Creek Cycles 8 and 9 as well as for the Hatch I Cycle 1 core. In each of the two Oyster Creek cycles twelve statepoints were calculated. With the exception of a few statepoints in Cycle 9, the core power was at or near the rated level. The mean k-effective for both Oyster Creek cycles was 0.986 with a standard deviation of about 0.2%.
Calculations of seventeen statepoints spanning the entire length of Hatch I Cycle 1 resulted in an average k-effective of 0.985 with a standard deviation I of about 0.5%. The larger standard deviation of the Hatch k-effective may be l due to the plugging of the lower grid plate.
It is seen that in both the Oyster Creek and Hatch verifications N0DE-B underpredicts the core reactivity by about 1.5%, with a standard deviation of about 0.5%,
i Power Distribution Uncertainties l
A measure of the accuracy of the calculated power distribution is provided by the comparison of measured TIP distributions with N0DE-B-predicted TIPS. GPU Nuclear's comparisons of these data were made for each of the 12 statepoints spanning Cycle 8 and again for each of the twelve statepoints spanning Cycle
- 9. In addition to these Oyster Creek comparisons, the N0DE-B model was verified against TIP and garrna scan measurements from Hatch 1 Cycle 1.
The verification from the Oyster Creek TTP data leads to a nodal uncertainty of 7.65%. Verification of the N0DE-B model against the Hatch 1 Cycle 1 measured TIP data results in a nodal uncertainty of 9.14%. Comparisons with Hatch 1 end-of-cycle 1 gamma scan measurements yield a nodal uncertainty of 7.95%. These results indicate that based on a data base derived from the operation of two Oyster Creek cycles and one Haten cycle, N0DE-B calculates nodal power distributions to within 9.14%. It is expected, therefore, that in i core related analyses involving nodal powers, GPU Nuclear will include an uncertainty of 9.14%.
3.3 Methodology Uncertainties In order to test the validity of the NODE-B model, operating data from two reactors were used including about forty operating states. These states provide an adequate data base for determining NODE-B uncertainties in predicting power distributions and hot and cold reactivities. Based on the calculation-to-measurement comparisons for these states, it is concluded that GPU Nuclear N0DE-B predictions of cold reactivity are accurate to within 0.5%
with a standard deviation of 0.3%, the hot reactivity predictions are accurate to within 1.5% with a standard deviation of 0.5%, and the nodal power predictions are accurate to within 9%.
Based on the review of the MODE-B methodology and on the verification of the code's ability to reproduce measured data, it is concluded that the code represents an acceptable methodology for performing three-dimensional steady state BWP reload calculations for the Oyster Creek core, that suitable {
comparisons to operating data were made and that there is a satisfactory agreement between the calculation results and the measurements, and that GPUN has therefore demonstrated an acceptable ability to use the code in cases in which the fuel loading and operating conditions are similar to those of Oyster Creek Cycles 8 and 9.
4.0 CONCLUSION
S The staff, with the assistance of consultants from Brookhaven National Laboratory, has reviewed the GPUN topical report TR-021, Revision 0, submitted by GPUN to describe and justify the methodology to be used in licensing calculations involving steady state BWR core characteristics. The review evaluated the methodology and the ability of GPUN to use the methodology.
Rased on this review we conclude that the CPM code as used by GPUM is acceptable for applicable BWR licensing calculations.
5.0 REFERENCES
- 1. Letter from R. F. Wilson, GPU Nuclear, to J. A. Zwolinski, NRC, March 25, 1986, "0yster Creek ... . Reload Topical Report."
- 2. EPRI-NODE-B, Advanced Recycle Methodology Program System Documentation Part II, Chapter 17, Rev. O September 1977.
- 3. EPRI-THERM-B, Advanced Recycle Methodology Program System Documentation Part II, Chapter 17, Rev. O, September 1977.
- 4. ARMP: Advanced Recycle Methodology Program Systems Documentation, Part I, Chapter 1 Rev. O, September 1977.
- 5. BWR Hybrid Power Shape Monitoring System EPRI NP 3195-CCM Yolumes 1 and 2, February 1981 and Volume 3, September 1983.
- 6. D. L. Delp, D. L. Fischer, J. W. Harriman, and M. J. Steadwell, " FLARE -
A Three-Dimensional Boiling Water Reactor Simulator," GEAP-4598, July 1964.
- 7. G. S. Lellouche and B. A. Zolotar, " Mechanistic Model for Predicting Two-Phase Void Fraction for Water in Vertical Tubes, Channels and Rod Bundles," EPRI-NP-2246-SR, February 1982.
- 8. Letter from R. F. Wilson, GPU Nuclear, to NRC, April 16, 1987, "0yster Creek .... Reload Topical Report 021."
- 9. Letter from R. F. Wilson, GPU Nuclear, to NRC, July 10, 1987, "0yster Creek . .. . Reload Topical Report 021."
Principal Contributor: H. J. Richings Date: September 22, 1987
.-