ML19283B612
| ML19283B612 | |
| Person / Time | |
|---|---|
| Issue date: | 02/12/1979 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19283B611 | List: |
| References | |
| NUDOCS 7903050064 | |
| Download: ML19283B612 (19) | |
Text
-
TOPICAL REPORT EVALVATI0ff CENPD-252-P.:
" Blowdown Analysjs itethod_- Method for the Analysis _of Blowdown Induced Forces in a Reactor Vessel."
A.
Summary of_Topi, cal Report This report describes the methods and procedures that are utilized by Combustion Engineering in calculating hydrodynamic loadings on a reactor coolant system undergoing a postulated loss-of-coolant accident (LOCA).
The hydrodynamics for both the subcooled and the saturated portions of the blowdown are calculated with the CEFLASH-4B digital computer program.
A post-processing routine for the CEFLASH-4B program is used to calculate the hydraulic forces acting on the reactor vessel internals as a result of momentum changes in the coolant fluid.
At present, a rigid boundary assumption is made in the hydraulic analyses. The fluid boundaries are assumed to be constant and at rest during the CE-FLASH-4B analysis.
The CEFLASH-4B computer program (References 1, 2 and 3) solves the con-servation equations for mass and energy, the one-dimensionil continuity equation, and the equation of state for water. The CEFLASH-48 program permits the user to select the nodal representation that results in the best finite differencing of the fluid system to be analyzed.
The program then solves the conservation equations for each node and the one-dimensional momentum equation for each flow path between nodes. CEFLASH-4B uses 7 9 0 3 0 5 0 0 @ Coi
_2-explicit solution techniques. Various options as well as user input parameters enable the program to nodel the reactor core, reactor coolant pumps, steam generators, and connesting piping in any con-figuration and operating model desired.
The thermodynamic and transport fluid properties in CEFLASH-4B are obtained from functional fits to the properties based on the 1967 ASME Steam Tables.
CEFLASH-4B is used to compute the pressure response of a system during a decompression transient. The transient pressure response can then be used to evaluate the system's everall dynamic structure response. The asymmetric pressure field in the downcomer annulus of a PWR can be obtained. This pressure field can then be integrated over the core support barrel area to obtain the total dynamic load on the core support barrel.
The analysis is performed for the subcooled decompression and early saturation periods of the transient, where the hydraulic loads are greatest.
These loads are used for the structural evaluation of the reactor pressure vessel system, in conjunction with other loads associated with a postulated LOCA and with a safe shutdown earthquake. The capability of the CEFLASH-48 analysis to account for the acoustic wave phenomena induced in subcooled water is demonstrated.
I:
B.
Regulatory Evaluation CEFLASH-4B is a generalized computer program; however, the review of the CEFLASH-4B program was limited to th'd' representation of a PWR primary coolant system subjected to a postulated loss-of-coolant accident (LOCA), specifically during the subcooled and early saturation periods of the decompression transient.
The NRC evaluation of the CEFLASH-4B program covered three maJur areas of review:
1.
Analytical Development 2.
Application and System Modeling 3.
Code Verification The application of the one-dimensional CEFLASH-4B program to the calcula-tion of the multi-dimensional downcomer pressure field represents the major portion of this review.
1.0 Conservation Eauations The basic assumption of equilibrium of the steam / water mixture is employed in the CEFLASH-4B program.
The subcooled, transition, and early saturated regimes, for which the equilibrium fluid assumption is acceptable, can be evaluated. It is during this portion of the blowdown that the LOCA hydraulic loads are greatest.
The mass, momentum and energy conservation equations are developed for a one-dimensional system. Density and mass flow are assumed to be uniform over the space and solution time interval. Friction and buoyancy are included in the momentum equation.
Included in the energy
. equation is heat transfer associated with a reactor core and with a steam generator.
2.0 EquationofStateand,SonicVelogi,g'"
The equation of state used in CEFLASH dB is a set of functional fits to the data for liquid and steam properties for single and two-phase conditions. The properties values obtained from these fits were checked against the 1967 ASME Steam Tables and show good agreement over the range of interest.
The sonic velocity in water is not explicitly used in the CEFLASH-4B analysis.
It can be derived, or implied, from the properties fits for an isentropic expansion process. The implied sound speed has been demonstrated to be in good agreement with the ASME values over the range of interest.
3.0 Method of Solution The solution of the conservation equations, for the analyse of hydro-dynamic loadings, is obtained by using the explicit numerical solution technique. The solution time step for stability is based on the system noding and the local acoustic wave speed. A time step study was per-formed for a typical LOCA loads model and the time step selected for licensing calculations is one-half that of the largest time step studied which showed a variation in the solution convergence.
The solution is assumed to be converged for the LOCA loads calculation when the time history results of the predicted pressures and momentum
. flow parameters are all nearly identical to the respective values from the preceeding smaller time step trial.
4.0 Discharge Flow Model and Non-Equilib_rium Effects The discharge flow model for the postulated break is the system forcing function. As such, the treatment of the subcooled critical flow and potential non-equilibrium effects must be properly accounted for in the development of the discharge flow model.
Comoustion Engineering uses the CE critical flow model for computing the subcooled and saturated critical fluid discharge at the break. The CE correlation accounts for the non-equilibrium nature of the critical discharge during the subcooled portion of the transient.
The CE critical flow model also yields higher, on the order of 10%, loads across the internal components than the equilibrium critical flow formulations.
The Moody and the Homogeneous Equilibrium models were used for this comparison.
The break characteristics are modeled in CEFLASH-4B as a function of location in the coolant system, total area and time to develop the total area. The guillotine offset area is properly modeled as the two ends of the pipe separate to the full double-ended area or partial offset area.
. ~The opening tine history of a break in the NSS primary piping system is plant specific, and it depends on considerations such as the location of the break, the stiffness and mass of the piping system involved, and the type and location of pipe restraints / supports being used.
Combustion Engineering utilizes a mechanistic approach based on non-linear structural analysis techniques and the conservative assumption of instantaneous crack propagation to determine realistic break opening times. The break opening schedules, for the System 80 generic class of plants, are documented in Reference 4.
The hydrodynamic loads are directly proportional to the differential pressures applied to the structure.
If, as a result of non-equilibrium the pressure at the break plane falls below the saturation pressure of the fluid, prior to the development of two-phase fluid conditions, an increased loading could result.
Tne effects of non-equilibrium could also increase the loading on the core support cylinder structure if this pressure "undershoot" could be transmitted to the downcomer annulus.
The reactor coolant system depressurization rate is, in part, a function of the break opening time and, in part, a function of the break area.
Combustion Engineering has assessed the potential for non-equilibrium effects to occur in a PWR blowdown as well as assessing the resulting affects should non-equilibrium conditions be present. The behavior of experiments, with initial conditions and break characteristics similar to the Combustion Engineering nechanistic break, were studied. For these conditions non-equilibrium effects were not observed. An assessment of the affects of a pressure "undershoot" on the impulse loading of the core support-barrel was performed based on the analysis performed in reference 5.
For this extreme case, the impulse loading increased less than one percent.
5.0 Multi-DimensionalRegion_Modelg The downcomer region of a PWR can be considcred at least 2-dimensional for analytical evaluations of subcoo'ed decompression transients and, as such, should be properly modeled. Since CEFLASH-48 solves the 1-dimensional conservation equations, it was necessary to investigate the modeling procedures used to represent the multi-dimensional aspects of this region.
The modeling techniques used to represent a physical system can affect the results of the calculation. Not only must the mathematical equa-tions be stable and the solution converged, but the nodal network representation should not exert undue influence (i.e., non-conservatism) on the calculation.
A set of experimental test geometries was selected to tudy the effects of the modeling techniques on the results of the calculations, by comparing the Combustion Engineering calculations to thnse performed by the NRC. The selected geometries contain two important features typical of PWRs: a downcomer annulus region and a core simulator region.
While the systems selected may not be properly scaled to a PWR system, in terms of the ratio of the downcomer length to circumferential length, for the analysis of a subcooled decompression transient, the system beh'avior should be well defined for the postulated transients.
If the system is properly modeled, the an,alysis will predict the expected transient behavior. The selected tests were instrumented to obtain subcooled decompression data. The selected geonetries were:
(1) LOFT Test L1-2.
This test was designed to represent a PWR during decompression.
The break was designed to represent a large, inlet nozzle rupture.
(2) Containment Systems Experiment (CSE) Test B-75.
This test was run at an initial pressure of 1000 psig and the break was designed to represent a large, inlet nozzle rupture.
In addition to the experimental test geometries, three problens using simplified geometries were developed to evaluate the multi-dimensional nodal approach. The results of these analyses were compared with multi-dimensional computer code calculations (Reference 6).
The geometries represented typical PWR regions, such as the nozzle to downcomer interface region, flow obstructions, and the un-wrapped downcomer region.
In addition to the multi-dimensional code calcu-lations, similar calculations using RELAP4/ MODS were performed for comparison with the CEFLASH-48 results (Reference 7).
The magnitude of the pressure wave penetrating into the downcomer annulus is the forcing function which determines the resulting hydraulic
-9 loading on the vessel internals. The inertial factor used in the CEFLASH-4B computer code to represent the flow path between the inlet nozzle and the downcomer is calculated as the sun of the inertial factors within the nozzle and inertial factor within the downconer.
One of the above simple problems was developed to study the effects of modeling at this location.
Comparisons of the CEFLASH-4B analyses for the three sample problems to the analyses performed by the NRC demonstrate the capability of the CEFLASH-4B computer program to account fc" the acoustic wave transmission and reflection phenomenon induced in subcooled water. The CEFLASH-4B nodal representation for multi-dimensional regions is shown to be acceptable within the computer program limitations which restrict the amount of spacial detail available.
The methodology and nodal representations are further justified by the good agreement obtained for the CEFLASH-4B analyses of the select :d test geometries.
6.0 Fluid-Structure Coupling Fluid-structure interaction is not included in the CEFLASH-4B analysis of subcooled blowdown. Fluid boundaries are assumed to be rigid and at rest.
7.0 NRC Audit Calculations The NRC has performed independent audit calculations for a System 80 PWR.
In addition, calculations for the selected test geometries were performed.
The WHAM /M00-007 computer code was used to perform the NRC audit calculations (Reference 8). The modeling techniques employed for these calculations were based on the evaluations of the three simple problems usedtostudythenetworkmodelingapproach'lReference9and10).
C.
Regulatory Position The CEFLASH-4B computer program is used by Combustiong Engineering to evaluate the hydrodynamic loadings on the reactor coolant system following a postulated loss-of-coolant accident (LOCA).
Fluid-structure interaction is not included in the CEFLASH-4B analysis of subcooled blowdown. Fluid boundaries are assumed to be rigid and at rest. The downcomer region of the PWR is modeled to allow for the calculation of the induced hydro-dynamic loads on the core support barrel. The methodology used results in a conservative calculation of the induced hydrodynamic loads on the reactor coolant system, reactor ves el supports, and reactor internals following a postulated LOCA.
1.0 Evaluation of Analytica. Methods The solution of the conservation equations and the eqt ation of state in the CEFLASH-4B computer program can be shown to be nearly equivalent to the vector momentum equation governing nearly incompressible, low speed flow for a multi-dimensional analyses. Therefore, the use of the CEFLASH-4B code to evaluate subcooled decompression transients, for the multi-aimensional nodal models, is scceptable.
.. The potential effects of non-equilibrium have been addressed in an acceptable manner, and changes to account explicitly for this phencnenon are not required for licensing calculations.
The discharge flow model used for the analysis of subcooled decompres-sion transients, the CE critical flow model, has been shown to yield con-servative results for the calculation of the hydrodynamic loads.
CEFLASH-4B is a generalized computer program for the analysis of thermal-hydraulic systems, and the user has a number of options available for an ana-lysis. During the course of this review a number of these options were ex-plored and subsequently accepted for a licensing calculation. These options are listed in the Summary section to this report and are also identified in the audit analyses.
2.0 Break Characteristics The break area models used in the CEFLASH-4B computer program are acceptable.
The break opening time history of a break in the NSS prinary piping system is plant specific, and it depends on con-derations such as the location of the break, the stiffness and mass of the piping system involved, and the type and location of pipe restraints / supports being used. Combustion Engineering utilizes a mechanistic approach based on non-linear structural analysis techniques and the conservative assumption of instantaneous crack propagation to determine realistic break opening times.
- 3.0 System Modeling and Code Verification The application of the CEFLASH-4B conputer program for the analyses of multi-dimensional fluid transients can be shown to,b,e equivalent to the vector momentum equation for nearly incompressible, low speed flow. This is demonstrated by comparisons of the simplied geometry problems to multi-dimensional computer code results. Further verification is obtained by comparing the CEFLASH-4R analyses to the selected experimental test geometry data. The evaluation of the nodeling techniques employed by Combustion Engineering, for the representation of a PWR, is based on the results and observations made from the comoarative analyses perfomed for the selected test geometries, and for the simplied geometry problem calculations.
There are four fundamental areas associated with this evaluation for the nodal hydraulic representation of the systen:
1.
The modeling of the primary coolant loops, 2.
The modeling of the nozzle to downcomer interface, 3.
The representation of the multi-dimensional downcomer region
- and, 4.
The representation of the vessel internals; lower plenum, core, upper plenum.
The representation of the primary coolant loops as one-dimensiona: pipes is acceptable. The piping system can be represented, for practical purposes, as one-dimensional. When the proper engineering loss factors are accounted for, the one-dimensional flow equation is sol.ed correctly.
The method for modeling the nozzle to downcomer region is acceptable.
The inertial factor used in the CEFLASH-4B computer program is obtained by summing the inertial factors for the nozzle with the inertial factors for the downcomer to nozzle interface. This method properly accounts for the geometry of this region.
The procedures used by Combustion Engineering, to generate the nodal representation of a PWR for licensing calculations, result in a conser-vative hydraulic model.
This is, in part, a result of the hydraulic nodal etwork employed to represent the downcomer annulus region.
Independent audit calculations performed by the NRC show that the nodal network developed by Combustion Engineering for the LOCA analysis is conservative.
Computer limitations restrict the amount of detail that can be specified in the CEFLASH-4B model.
This limits the user's ability to select a nodal representation, or finite differencing of the fluid system to be analyzed.
Combustion Engineering performed a sensitivity study for the nodal repre-sentation of the downcomer annulua. Based on these studies, which in-cluded the evaluation of the hydraulic loads on the core support barrel, a design application model was selected.
This model is acceptable for licensing calculations.
This. model also includes the representations of the remainder of the primary system.
The steam generator, pressurizer, coolant pumps and the vessel internals are described using the node and flow path modeling technique.
The multi-dimensional region modeling has been verified by comparing the CEFLASH-48 results to 2-dimensional computer cude analyses of the three simplified geometry problem.
These com-parisons demonstrate the near equivalence of the CEFLASH-4B solutions for incompressible, low speed ficw conditions. Further verifications were made by comparisons of the CEFLASH-4B results to the selected test geometry data. The analyses of the LOFT L1-2 and the Containment Systems Experiment test B-75 showed good agreement with the test data and the analytical results compared favorably with the NRC calculations.
In general, the differential pressures obtained from the CEFLASH-4B analyses are larger than those reported in the test results and larger than those calculated with the NRC methodology.
These differential pressures are similar to those which would be used to obtain the integrated hydraulic loads on the PWR core suoport barrel.
The methodology employed by Combustion Engineering to generate the nodal representation of the PWR system, specifically at the nozzle to downcomer interface, and for the development of the downconer region itself, shall be incorporated into the final topical report documentation (see Appendix A).
The development of a model using ;his methodology is acceptable for licensing calculations.
4.0 Audit Calculations Conbustion Engineering has performed an audit analysis with both ahotlegnozzlebreakandapartialdoubie-endedguillotine rupture at the reactor inlet nozzle using the mechanistic break data.
In addition, an analysis was performed for the inlet nozzle rupture using the design model but with a full double-ended guillotine break.
Comparisons of the Combustion Engineering analyses to the NRC inde-pendent analyses shows that the CEFLASH-4B methodology predicts higher hydraulic loads, for all cases evaluated.
The model developed by Combustion Engineering is based on the methodology reviewed for this evaluation. The CEFLASH 4B calculation of a postu-lated LOCA in a PWR for the purpose of determining the resultant hydro-dynamic loads on the system, is acceptable for licensing calculations.
5.0 Fluid-Structure _ Coupling Fluid-structure interaction is not included in the CEFLASH-4B analysis of subcooled blowdown.
Fluid boundaries are assumed to be rigid and at rest.
Analytical evaluations of the effects of fluid-structure coupling, by a number of researchers (References 11, 12, and 13), have shown that for a coupled analysis the frequency and amplitude motions are lower than for an uncoupled analysis with the consequence of generally lower induced stresses.
~
_ The rig'id boundary assumption used in the CEFLASH-4B analysis is acceptable for licensing calculations. However, the NRC will continue to monitor the on-going research in this area and will require, if necessary, a re-evaluation of the rigid boundary assumption. Part of this research will be performed in the German HDR subcooled blowdown experiments (Reference 14). Combustion Engineering has committed to perform a pre-test analysis of this experiment when test conditions have been established (Reference 15).
D.
Summary The Combustion Engineering CEFLASH-4B computer program is used to evaluate the subcooled decompression and early saturation response of a PWR primary coolant system following a postulated loss-of-coolant accident. This topical report describes the assumptions used and the methodology employed by Combustion Engineering to perform this evaluation.
Subject to the limitations of this review, the CEFLASH-48 computer p ogram is an acceptable code for evaluating the subcooled decompression response of a PWR primary coolant system following a postulated loss-of-coolant accident. These limitations are:
(1) The CE critical flow model is to be used.
(2) The break opening schedules, including location, size and time based on the mechanistic break model employed by Combustion Engineering are to be referenced for licensing calculations.
(3) The Combustion Engineering design model for the annulus repre-sentation is to be used for licensing calculations.
. {4) The evaluation of the blowdown induced forces following a postulated LOCA is acceptable provided a CEFLASH-4A licensing calculation is perfonr:d to obtain the hydraulic input data.
The responses to the NRC requests for additional information, as listed in Appendix A of this evaluation, are to be incorporated into the final version of the topical report to describe and develop the Combustion Engineering modeling nethodology and to identify the assumptions and acceptable input parameters for a licensing calculation.
Combustion Engineering has committed to perform a pre-test analysis of the German HDR subcooled blowdown experiment when test conditions becomes available.
REFERENCES
'1.
CENPD-252-P," Blowdown Analysis t'ethod, Method for the Analysis of Blowdown Induced Forces in A Reactor Vessel," Combustion Enginer-ing, Reactor Design, December, 1977.
2.
CENPD-252-P, Amendment 1-P, Combustion engineering, Blowdown Loads Group, August, 1978.
3.
CENPD-252-P, Amendment 2-P, Combustion Engineering, Blowdown Loads Group, October, 1978.
4.
CENPD-168-A, " Design Basis Pipe Breaks for the Combustion Engineering Two-Loop Reactor Coolant System," Combustion Engineering, June 1977.
5.
Jackson, J.F., " Nuclear Reactor Safety Quarterly Progress Report:
January 1 - March 31, 1977," LA-NUREG-6842-PR, June, 1977.
6.
Cole, R.K., " Analysis of NRC Sample Problems Using the CSQ Computer Program," SANDIA, (to be published).
'7.
Cole, R.K., " Analysis of NRC Sample Problems Using the RELAP4/M005 Computer Program," SANDIA (to be published).
8.
Throm, E.D., " Availability of the WHAM / MOD-007 Subcooled Blowdown Hydraulic Analysis Code," memo to Z.R. Rosztoczy, Chief, Analysis Branch, DSS, May 13, 1976.
9.
Cole, R.K., "An Analysis of the Method of Streeter and Wylie for Multidimensional Wave Propagation," SAND 77-0549, Sandia Laboratories, July, 1977.
- 10. Berman, M., Cole, R.K., et.al., "LOCA Analysis Quarterly Report April-June,1977, SAND 77-1530, Sandia Laboratories, December, 1977 (Section II).
- 11. Prelewicz, D.A., " Hydraulic Pressure Pulses with Structural Flexibility:
Test and Analysis, "WAPD-TM-1227, Bettis Atomic Pcwer Laboratory, April, 1976.
- 12. Takeuchi, K., " Hydraulic Force Calculation with Hydrostructural Interactions," Nuclear Technology, Vol. 39, July,1978, pages 155-166.
- 13. Rivard, W.C., and M.D. Torrey, " Fluid-Structure Response of a Pressurized Water Reactor Core Barrel During Blowdown," NUREG/CR-0264, LA-7407, September, 1978.
- 14. Krieg, R., E.G. Schlechtendahl, K.- H. Scholl, " Design of the HDR Experimental Program on Blowdown Loading and Dynamic Response of PWR-Vessel Internals," Nuclear Engineering and Design, Vol. 43, 1977.
- 15. Letter from A.E. Scherer, Licensing Manager, Combustion Engineering, to K. Kniel, Chief, LWR-2, Division of Project Management, LD-77-032, April 7, 1977.
APPEi: DIX A The responses to the following NRC requests for additional information (reference NRC letter from D.B. Vassallo, DPM, to A.E. Scherer, Licensing Manager, dated May 10,1978) are to be incorporated into the final version of the report to describe and develop the Combustion Engineering nodeling methodology and to identify the assumptions and acceptable input parameters for a licensing calculation:
questions 1.1 2.1 3.1 1.2 2.2 3.2 1.3 2.3 1.4 2.4 without listings 2.5j 2.6
-