Regulatory Guide 1.157

(Task Rs 701-4), Best-Estimate Calculations of Emergency Core Cooling System Performance

ML003739584
Person / Time
Issue date:	05/31/1989
From:	Office of Nuclear Regulatory Research
To:
References
RG-1.157
Download: ML003739584 (20)

U.S. NUCLEAR REGULATORY

COMMISSION

REGULATORY

May 1989 GUIDE OFFICE OF NUCLEAR REGULATORY

RESEARCH REGULATORY

GUIDE 1.157 (Task RS 701-4) BEST-ESTIMATE

CALCULATIONS

OF EMERGENCY

CORE COOLING SYSTEM PERFORMANCE

A. INTRODUCTION

Section 50.46, "Acceptance Criteria for Emer gency Core Cooling Systems for Light-Water Nuclear Power Reactors," of 10 CFR Part 50, "Domestic Li censing of Production and Utilization Facilities," re quires that light-water nuclear reactors fueled with uranium oxide pellets within cylindrical zircaloy clad ding be provided with emergency core cooling systems (ECCS) that are designed in such a way that their calculated core cooling performance after a postu lated loss-of-coolant accident (LOCA) conforms to certain criteria specified in paragraph

50.46(b).

Para graph 50.46(b)(1)

requires that the calculated maxi mum temperature of fuel element cladding not be ~ greater than 2200'F. In addition, paragraphs

50.46(b) (2) through (b) (5), which contain required limits for calculated maximum cladding oxidation and maximum hydrogen generation, require that calcu lated changes in core geometry remain amenable to cooling and that long-term decay heat removal be provided.

On September

16, 1988, the NRC staff amended the requirements of § 50.46 and Appendix K, "ECCS Evaluation Models" (53 FR 35996), so that these regulations reflect the improved understand ing of ECCS performance during reactor transients that was obtained through the extensive research performed since the promulgation of the original requirements in January 197

4. Paragraph

50.46(a)(1)

now permits licensees or applicants to USNRC REGUIATORY

GUIDES Regulatory Guides are Issued to describe and make available to the pub lic methods acceptable to the NRC staff of implementing specific parts of the Commission's regulations, to delineate techniques used by the staff in evaluating specific problems or postulated accidents, or to pro vide guidance to applicants.

Regulatory Guides are not substitutes for regulations, and compliance with them is not required.

Methods and solutions different from those set out In the guides will be acceptable If they provide a basis for the findings requisite to the issuance or continu ance of a permit or license by the Commission.

This guide was issued after consideration of comments received from the public. Comments and suggestions for Improvements In these guides are encouraged at all times, and guides will be revised, as ap propriate, to accommodate comments and to reflect new Information or experience.

Written comments may be submitted to the Regulatory Publications Branch, DFIPS, ARM, U.S. Nuclear Regulatory Commission, Washing ton, DC 20555.use either Appendix K features or a realistic 1 evaluation model. These realistic evaluation models 2 must include sufficient supporting justification to demonstrate that the analytic techniques employed realistically describe the behavior of the reactor system during a postulated loss-of-coolant accident.

Paragraph

50.46(a)(1)

also requires that the uncertainty in the realistic evaluation model be quantified and considered when comparing the results of the calculations with the applicable limits in paragraph

50.46(b) so that there is a high probability that the criteria will not be exceeded.

This regulatory guide describes models, 3 correla tions, 4 data, model evaluation procedures, and meth ods that are acceptable-to the NRC staff for meeting the requirements for a realistic or best-estimate calcu lation of ECCS performance during a loss-of-coolant accident and for estimating the uncertainty in that 1 For the purpose of this guide, the terms "best-estimate" and "realis tic" have the same meaning. Both terms are used to indicate that the techniques attempt to predict realistic reactor system thermal-hydraulic response.

Best-estimate is not used in a statistical sense in this guide. 2 IThe term "evaluation model" refers to a nuclear plant system com puter code or any other analysis tool designed to predict the aggregate behavior of a reactor during a loss-of-coolant accident.

It can be either best-estimate or conservative and may contain many correlations or models. MIhe term "model" refers to a set of equations derived from funda mental physical laws that is designed to predict the details of a specific phenomenon.

4 The term "correlation" refers to an equation having empirically de termined constants such that it can predict some details of a specific phe nomenon for a limited range of conditions.

The guides are issued In the following ten broad divisions:

1. Power Reactors 2. Research and Test Reactors 3. Fuels and Materials Facilities

4. Environmental and Siting 5. Materials and Plant Protection

6. Products

7. Transportation

8. Occupational Health 9. Antitrust and Financial Review 10. General Copies of issued guides may be purchased from the Government Printing Office at the current GPO price. Information on current GPO prices may be obtained by contacting the Superintendent of Documents, U.S. Government Printing Office, Post Office Box 37082, Washington, DC 20013-7082, telephone

(202)275-2060

or (202)275-2171.

Issued guides may also be purchased from the National Technical Infor mation Service on a standing order basis. Details on this service may be obtained by writing NTIS, 5285 Port Royal Road, Springfield, VA 22161.

calculation.

Methods for including the uncertainty in the comparisons of the calculational results to the cri teria of paragraph

50.46(b), in order to meet the re quirement that there be a high probability that the criteria would not be exceeded, are also described in this regulatory guid

e. Paragraph

50.46(a) also per mits licensees to use evaluation models developed in conformance with Appendix K. Other models, data, model evaluation proce dures, and methods will be considered if they are supported by appropriate experimental data and technical justification.

Any models, data, model evaluation procedures, and methods listed as accept able in this regulatory guide are acceptable in a ge neric sense only and would still have to be justified to the NRC staff as being appropriately applied and ap plicable for particular plant applications.

The regulatory position in this regulatory guide lists models, correlations, data, and model evaluation procedures that the NRC staff considers acceptable for realistic calculations of ECCS performance.

It also provides a description of the acceptable features of best-estimate computer codes and acceptable methods for determining the uncertainty in the calcu lations.

The Advisory Committee on Reactor Safeguards has been consulted concerning this guide and has concurred in the regulatory position.

Any information collection activities discussed in this regulatory guide are contained as requirements in 10 CFR Part 50, which provides the regulatory basis for this guide. The information collection require ments in 10 CFR Part 50 have been cleared under OMB Clearance No. 3150-0011.

B. DISCUSSION

The criteria set forth in § 50.46, "Acceptance Criteria for Emergency Core Cooling Systems for Light-Water Nuclear Power Reactors," and the cal culational methods specified in Appendix K were promulgated in January 1974 after extensive rulemaking hearings and were based on the under standing of ECCS performance available at that time. In the years following the promulgation of those rules, the NRC, the nuclear industry, and several for eign institutions have conducted an extensive pro gram of research that has greatly improved the un derstanding of ECCS performance during a postulated loss-of-coolant accident.

The methods specified in Appendix K were found to be highly con servative;

that is, the fuel cladding temperatures ex pected during a loss-of-coolant accident would be much lower than the temperatures calculated using Appendix K methods. In addition to showing that Appendix K is conservative, the ECCS research pro-vided information that allows for quantification of that conservatism.

The results of experiments, com puter code development, and code assessment allow more accurate calculations, along with reasonable es timates of uncertainty, of ECCS performance during a postulated loss-of-coolant accident than is possible using the Appendix K procedures.

It was also found that some plants were being re stricted in operating flexibility by limits resulting from conservative Appendix K requirements.,Based on the research performed, it was determined that these re strictions could be relaxed through the use of more realistic calculations without adversely affecting safety. The Appendix K requirements tended to di vert both NRC and industry resources from matters that are relevant to reactor safety to analyses with as sumptions known to be nonphysical.

In recognition of the known conservatisms in Ap pendix K, the NRC adopted an interim approach in 1983, described in SECY-83-472,5 to accommodate industry requests for improved evaluation models for the purpose of reducing reactor operating restrictions.

This interim approach was a step in the direction of basing licensing decisions on realistic calculations of plant behavior.

Although the approach permits many "best-estimate" methods and models to be used for licensee submittals, it retains those features of Ap pendix K that are legal requirements.

The current revision of § 50.46 permits ECCS evaluation models to be fully "best-estimate" and re moves the arbitrary conservatisms contained in the required features of Appendix K for those licensees wishing to use these improved methods. Safety is best served when decisions concerning the limits within which nuclear reactors are permitted to operate are based upon realistic calculations.

This approach is currently being used in the resolution of almost all reactor safety issues (e.g., anticipated transients with out scram, pressurized thermal shock, and operator guidelines)

and is now available for one of the last remaining major issues still treated in a prescriptive manner, the loss-of-coolant accident.

The NRC staff amended § 50.46 of 10 CFR Part 50 to allow realistic methods to be used for the ECCS performance calculations in place of the evaluation models that use the required Appendix K features.

This rule change also requires analysis of the uncer tainty of the best-estimate calculations and requires that this uncertainty be considered when comparing the results of the calculations to the limits of para graph 50.46(b) so that there is a high probability that " 6 Information Report from William J. Dircks to the Commissioners, dated November 17,1983, "Emergency Core Cooling System Analy sis Methods," SECY-83-472.

Available for inspection or copying for a fee in the NRC Public Document Room, 2120 L Street NW., Washington, DC.1.157-2 the criteria will not be exceeded.

In this manner, more realistic calculations are available for regulatory decisions, yet an appropriate degree of conservatism would be maintained.

Many of the methods and models needed for a S best-estimate calculation are the same as those used previously for evaluation model analyses.

Although licensees and applicants are well acquainted with them, explicit guidance on acceptable methods and models (based on NRC experience with its own best estimate advanced codes such as TRAC-PWR, TRAC-BWR, RELAP5, COBRA, and FRAP) would be useful. Further, the NRC has not previously pub lished acceptable methods for uncertainty analyses.

Therefore, guidance on methods acceptable to the NRC staff for calculating ECCS performance and for estimating the uncertainty are provided in the follow ing Regulatory Position.

C. REGULATORY

POSITION 1. BEST-ESTIMATE

CALCULATIONS

A best-estimate calculation uses modeling that attempts to realistically describe the physical proc esses occurring in a nuclear reactor. There is no unique approach to the extremely complex modeling of these processes.

The NRC has developed and as sessed several best-estimate advanced thermal hydraulic transient codes. These include TRAC PWR, TRAC-BWR, RELAP5, COBRA, and the .FRAP series of codes (References

1 through 6). These codes reasonably predict the major phenom ena observed over a broad range of thermal hydraulic and fuel tests. Licensees and applicants may use, but are not limited to, these codes and the specific models within them to perform best-estimate calculations of emergency core cooling system (ECCS) performance.

Since the NRC staff has not performed the plant-specific uncertainty analysis re quired by the revised § 50.46 of 10 CFR Part 50, the licensee must demonstrate that the code and models used are acceptable and applicable to the specific fa cility over the intended operating range and must quantify the uncertainty in the specific application.

General attributes expected in a best-estimate calcu lation are described here in Regulatory Position 1; special considerations for thermal-hydraulic best estimate codes are presented in Regulatory Position 2; and specific examples of features that are consid ered acceptable best-estimate models are given in Regulatory Position 3. Other models or correlations will be considered acceptable if their technical basis is demonstrated with appropriate data and analysis.

A best-estimate model should provide a realistic calculation of the important parameters associated with a particular phenomenon to the degree practical with the currently available data and knowledge of the phenomenon.

The model should be compared with applicable experimental data and should predict the mean of the data, rather than providing a bound to the data. The effects of all important variables should be considered.

If it is not possible or practical to consider a particular phenomenon, the effect of ignoring this phenomenon should not normally be treated by including a bias in the analysis directly, but should be included as part of the model uncertainty.

The importance of neglecting a particular phenome non should be considered within the overall calcula tional uncertainty.

Careful consideration should be given to the range of applicability of a model when used in a best estimate code. When comparing the model to data, judgments on the applicability of the data to the situ ation that would actually occur in a reactor should be made. Correlations generally should not be extrapo lated beyond the range over which they were devel oped or assessed.

If the model is to be extrapolated beyond the conditions for which valid data compari sons have been made, judgments should be made as to the effect of this extrapolation and the effect should be accounted for in the uncertainty evalu ation. The fundamental laws of physics, well established data bases (e.g., steam tables), and sensi tivity studies should be used to assist in estimating the uncertainty that results from extrapolation.

A best-estimate code contains all the models nec essary to predict the important phenomena that might occur during a loss-of-coolant accident.

Best-estimate code calculations should be compared with applicable experimental data (e.g., separate-effects tests and in tegral simulations of loss-of-coolant accidents)

to de termine the overall uncertainty and biases of the cal culation.

In addition to providing input to the uncertainty evaluation, integral simulation data com parisons should be used to ensure that important phe nomena that are expected to occur during a loss-of coolant accident are adequately predicted.

This is an idealized characterization of a best-estimate code. In practice, best-estimate codes may contain certain models that are simplified or that contain conserva tism to some degree. This conservatism may be intro duced for the following reasons: 1. The model simplification or conservatism has little effect on the result, and therefore the development of a better model is not justified.

2. The uncertainty of a particular model is diffi cult to determine, and only an upper bound can be determined.

3. The particular application does not require a totally best-estimate calculation, so a bias in the calculation is acceptable.

The introduction of conservative bias or simplifi cation in otherwise best-estimate codes should not, 1.157-3 however, result in calculations that are unrealistic, that do not include important phenomena, or that contain bias and uncertainty that cannot be bounded.

Therefore, any calculational procedure determined to be a best-estimate code in the context of this guide or for use under paragraph

50.46(a)(i)

should be compared with applicable experimental data to en sure that the calculation of important phenomena is realistic.

2. CONSIDERATIONS

FOR THERMAL HYDRAULIC

BEST-ESTIMATE

CODES Some features that are acceptable for use in best-estimate codes are described in the following paragraphs.

Models that address these features may be used with the basic proviso that a specific model is acceptable if it has been compared with applicable experimental data and shown to provide reasonable predictions.

Reference

7, "Compendium of ECCS Research for Realistic LOCA Analysis" (NUREG-1230), provides a summary of the large ex perimental data base available, upon which best estimate models may be based. While inclusion in Reference

7 does not guarantee that the data or model will be acceptable, the report describes and references a large body of data generally applicable to best-estimate models. NUREG-1230

also provides documentation of NRC studies of the effect of reac tor power increase on risk, background information on the ECCS rule, and a description of the methodol ogy developed by NRC for estimating thermal hydraulic transient code uncertainty.

For any models or correlations used in a best estimate code, sufficient justification must be pro vided to substantiate that the code performs ade quately for the classes of transients to which it applied. In general, the features of best-estimate thermal-hydraulic transient codes have uncertaintie associated with their use for predicting reactor syster response.

These uncertainties should be considered as part of the overall uncertainty analysis described in Regulatory Position 4. 2.1 Basic Structure of Codes 2.1.1 Numerical Methods A best-estimate code uses a numerical scheme for solving the equations used to predict the thermal hydraulic behavior of the reactor. The numerical scheme is, in itself, a complex process that can play an important role in the overall calculation.

Careful numerical modeling, sensitivity studies, and evalu ations of numerical error should be performed to en sure that the results of the calculations are represen tative of the models used in the code. Numerical simulations of complex problems, such as those con sidered here, treat the geometry of the reactor in an approximate manner, making use of discrete volumes or nodes to represent the system. Sensitivity studies and evaluations of the uncertainty introduced by noding should be performed.

Numerical methods treat time in a discrete manner, and the effect of time-step size should also be investigated.

2.1.2 Computational Models A best-estimate code typically contains equations for conservation of mass, energy, and momentum of the reactor coolant and noncondensible gases, if im portant (e.g., air, nitrogen).

Energy equations are also used to calculate the temperature distribution in reactor system structures and in the fuel rods. The required complexity of these equations will vary de pending on the phenomena that are to be calculated and- the required accuracy of the calculation.

NRC staff experience with its own best-estimate computer codes has indicated that separate flow fields for dif ferent fluid phases, or types, and calculation of non equilibrium between phases may be required to calcu late some important phenomena (e.g., countercurrent flow, reflood heat transfer)

to an ac ceptable accuracy.

The NRC staff has also deter mined that certain phenomena require that the equa tions be solved in multiple dimensions.

However, one-dimensional approximations to three dimensional phenomena will be considered accept able if those approximations are properly justified.

Other basic code features include equations of state and other material properties.

Sensitivity studies and comparisons to data should be performed to deter mine the importance of the simplifications used. 3. BEST-ESTIMATE

CODE FEATURES 3.1 Initial and Boundary Conditions and Equipment Availability The heat generated by the fuel during a loss-of coolant accident depends on the power level of the reactor at the time of the loss-of-coolant accident and on the history of operation.

The most limiting initial conditions expected over the life of the plant should be based on sensitivity studies. It is not necessary to assume initial conditions that could not occur in com bination.

For example, beginning-of-life peaking fac tors together with end-of-life decay heat do not re quire consideration.

Given the assumed initial conditions, relevant factors such as the actual total power, actual peaking factors, and actual fuel condi tions should be calculated in a best-estimate manner. The calculations performed should be represen tative of the spectrum of possible break sizes from the full double-ended break of the largest pipe to a size small enough that it can be shown that smaller breaks are of less consequence than those already consid ered. The analyses should also include the effects of longitudinal splits in the largest pipes, with the split area equal to twice the cross-sectional area of the pipe. The range of break sizes considered should be 1.157-4 sufficiently broad that -the system response as a func tion of break size is well enough defined so that inter polations between calculations, without considering unexpected behavior between the break sizes, may be made confidently.

Other boundary and initial conditions and equip ment availability should be based on plant technical specification limits. These other conditions include, but may not be limited to, availability and perform ance of equipment, automatic controls, and operator actions. Appendix A to 10 CFR Part 50 requires that a single failure be considered when analyzing safety system performance and that the analysis consider the effect of using only onsite power and only offsite power. 3.2 Sources of Heat During a Loss-of-Coolant Accident Models should account for the sources of heat discussed below and the distribution of heat production.

3.2.1 Initial Stored Energy of the Fuel The steady-state temperature distribution and stored energy in the fuel before the postulated acci dent should be calculated in a best-estimate manner for the assumed initial conditions, fuel conditions, and operating history. To accomplish this, the ther mal conductivity of the fuel pellets and the thermal conductance of the gap between the fuel pellet and the cladding should be evaluated.

Thermal conduc tivity of fuel is a function of temperature and is de graded by the presence of gases in crack voids be tween fuel fragments.

An acceptable model for thermal conductivity should be developed from the in-pile test results for fuel centerline and off-center temperatures, taking into account the conductivity of gases in crack voids. Thermal conductance of the fuel-cladding gap is a strong function of hot gap size and of the composi tion and pressure of the gases in the fuel rod. The calculation of hot gap size should take into account U0 2 or mixed-oxide fuel swelling, densification, creep, thermal expansion and fragment relocation, and cladding creep. Fuel swelling is a function of temperature and burnup. Fuel densification is a func tion of burnup, temperature, and initial density. Den sification can result from hydrostatic stresses imposed on fuel during pellet-cladding mechanical interaction and should be considered.

Fuel creep is a function of time, temperature, grain size, density, fission rate, oxygen-to-metal ratio, and external stress. Fuel ther mal expansion represents dimensional changes in unirradiated fuel pellets caused by changes in tem perature.

An acceptable model for the above fuel pa rameters should be based on in-pile and out-of-pile test data. Cladding creep introduces compressive creep strain in cladding during steady-state operation, reducing the gap between the fuel pellet and clad ding. Cladding creep is a function of fast neutron flux (>1 MeV), cladding temperature, hoop stress, and material.

Cladding materials may be cold worked and stress-relieved or fully recrystallized, and there is a significant difference in the magnitude of creepdown between these materials.

During pellet cladding mechanical interaction, cladding experi ences deformation from tensile creep, which is sig nificantly different from that caused by compressive creep. An acceptable model for cladding tensile creep should be based on in-reactor tensile creep data. Best-estimate fuel models will be considered ac ceptable provided the models include essential phe nomena identified above and provided their technical basis is demonstrated with appropriate data and analyses.

3.2.1.1 Model Evaluation Procedure for Stored Energy and Heat Transfer in Fuel Rods. A model to be used in ECCS evaluations to calculate internal fuel rod heat transfer should: a. Be checked against several sets of relevant data, and b. Recognize the effects of fuel burnup, fuel pellet cracking and relocation, cladding creep, and gas mixture conductivity.

The model described by Lanning (Ref. 8) com pared well to in-pile fuel temperature data. Best estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.2.1.2 Experimental Data for Stored Energy in Fuel Rods and Heat Transfer.

The correlations and data of Reference

9 are acceptable for calculating the initial stored energy of the fuel and subsequent heat transfer.

3.2.2 Fission Heat Fission heat should be included in the calculation and should be calculated using best-estimate reactiv ity and reactor kinetics calculations.

Shutdown reac tivities resulting from temperatures and voids should also be calculated in a best-estimate manner. The point kinetics formulation is considered an accept able best-estimate method for determining fission heat in safety calculations for loss-of-coolant acci dents. Other best-estimate models will be considered acceptable provided their technical basis is demon strated with appropriate data and analyses.

Control rod assembly insertion may be assumed if it is ex pected to occur. 3.2.3 Decay of Actinides The heat from radioactive decay of actinides, in cluding neptunium and plutonium generated during 1.157-5 operation as well as isotopes of uranium, should be calculated in accordance with fuel cycle history and known radioactive properties.

The actinide decay heat chosen should be appropriate for the facility's operating history. Best-estimate models will be con sidered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.2.4 Fission Product Decay Heat The heat generation rates from radioactive decay of fission products, including the effects of neutron capture, should be included in the calculation and should be calculated in a best-estimate manner. The energy release per fission (Q value) should also be calculated in a best-estimate manner. Best-estimate methods will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

The model in Reference

10 is consid ered acceptable for calculating fission product decay heat. 3.2.4.1 Model Evaluation Procedure for Fission Product Decay Heat. The values of mean energy per fission (Q) and the models for actinide decay heat should be checked against a set of relevant data. 3.2.5 Metal-Water Reaction Rate The rate of energy release, hydrogen generation, and cladding oxidation from the reaction of the zir caloy cladding with steam should be calculated in a best-estimate manner. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

For rods calculated to rupture their cladding during the loss-of-coolant accident, the oxidation of the in side of the cladding should be calculated in a best estimate manner. 3.2.5.1 Model Evaluation Procedure for Metal-Water Reaction Rate. Correlations to be used to calculate metal-water reaction rates at less than or equal to 1900'F should: a. Be checked against a set of relevant data, and b. Recognize the effects of steam pressure, pre oxidation of the cladding, deformation dur ing oxidation, and internal oxidation from both steam and U0 2 fuel. The data of Reference

11 are considered accept able for calculating the rates of energy release, hydro gen generation, and cladding oxidation for cladding temperatures greater than 1900 'F.3.2.6 Heat Transfer from Reactor Internals Heat transfer from piping, vessel walls, and inter nal hardware should be included in the calculation and should be calculated in a best-estimate manner. Heat transfer to channel boxes, control rods, guide tubes, and other in-core hardware should also be considered.

Models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.2.7 Primary to Secondary Heat Transfer (Not Applicable to Boiling Water Reactors)

Heat transferred between the primary and secon dary systems through the steam generators should be considered in the calculation and should be calcu lated in a best-estimate manner. Models will be con sidered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.3 Reactor Core Thermal/Physical Parameters

3.3.1 Thermal Parameters for Swelling and Rupture of the Cladding and Fuel Rods A calculation of thle swelling and rupture of the cladding resulting from the temperature distribution in the cladding and from the pressure difference be tween the inside and outside of the cladding, both as a function of time, should be included in the analysis and should be performed in a best-estimate manner. The degree of swelling and rupture should be taken into account in the calculation of gap conductance, cladding oxidation and embrittlement, hydrogen gen eration, and heat transfer and fluid flow outside of the cladding.

The calculation of fuel and cladding temperatures as a function of time should use values of gap conductance and other thermal parameters as functions of temperature and time. Best-estimate methods to calculate the swelling of the cladding should take into account spatially varying cladding temperatures, heating rates, anisotropic material properties, asymmetric deformation of cladding, and fuel rod thermal and mechanical parameters.

Best estimate methods will be considered acceptable pro vided their technical basis is demonstrated with ap propriate data and analyses.

3.3.2 Other Core Thermal Parameters As necessary and appropriate, physical and chemical changes in in-core materials (e.g., eutectic formation, phase change, or other phenomena caused by material interaction)

should be accounted for in the reactor core thermal analysis.

Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.1.157-6

3.4 Blowdown Phenomena

3.4.1 Break Characteristics and Flow In analyses of hypothetical loss-of-coolant acci dents, a spectrum of possible break sizes should be --' considered, as indicated in Regulatory Position 3.1. The discharge flow rate should be calculated with a critical flow rate model that considers the fluid conditions at the break location, upstream and down stream pressures, and break geometry.

The critical flow model should be justified by comparison to ap plicable experimental data over a range of conditions for which the model is applied. The model should be a best-estimate calculation, with uncertainty in the critical flow rate included as part of the uncertainty evaluation.

Best-estimate models will be considered acceptable provided their technical basis is demon strated with appropriate data and analyses.

3.4.1.1 Model Evaluation Procedure for Discharge Flow Rate. Critical flow models to be employed in ECCS evaluations should: a. Be checked against an acceptable set of rele vant data, b. Recognize thermal nonequilibrium conditions when the fluid is subcooled, and c. Provide a means of transition from nonequi librium to equilibrium conditions.

The uncertainties and bias of a correlation or model used to calculate critical flow should be stated, as well as their range of applicability.

The mechanistic thermal nonequilibrium and slip model of Richter (Ref. 12) compares well to small and large-scale test data (Ref. 13). 3.4.1.2 Experimental Data for Discharge Flow Rate. An acceptable set of relevant critical flow data should cover the fluid conditions, geometries, and types of breaks pertinent to light-water reactor loss-of-coolant accidents.

The following tests should be considered in establishing an acceptable set of relevant data: "* Marviken tests (Ref. 14) "* Moby Dick experiments (Ref. 15) "* Brookhaven critical flashing flows in nozzles (Ref. 16) "* Sozzi-Sutherland tests (Ref. 17) "* Edwards experiments (Ref. 18) "* Super Moby Dick experiments (Refs. 19 and 20)For critical flow from small breaks under strati fied conditions, currently acceptable test data for as sessing models and codes include those reported by: "* Anderson and Owca (Ref. 21) "* Reimann and Khan (Ref. 22) "* Schrock et al. (Refs. 23 and 24) 3.4.2 ECC Bypass The best-estimate code should contain a calcula tion of the amount of injected cooling water that by passes the vessel during the blowdown phase of the loss-of-coolant accident.

The calculation of ECC by pass should be a best-estimate calculation using analyses and comparisons with applicable experimen tal data. Although it is clear that the dominant proc esses governing ECC bypass are multidimensional, single-dimensional approximations justified through sufficient analysis and data may be acceptable.

Best estimate methods will be considered acceptable pro vided their technical basis is demonstrated with ap propriate data and analyses.

Cooling water that is not expelled, but remains in piping or is stored in parts of the vessel, should be calculated in a best-estimate manner based on applicable experimental data. 3.4.2.1 Model Evaluation Procedure for ECC Bypass. A correlation or model to be used to evaluate ECC bypass should: a. Be checked against an acceptable set of rele vant data, and b. Recognize the effects of pressure, liquid sub cooling, fluid conditions, hot walls, and sys tem geometry.

Uncertainties and bias in the correlations or models used to calculate ECC bypass should be stated, as well as the range of their applicability.

For scaled-down PWR downcomers, correlations by Beckner and Reyes (Ref. 25) compared well to the bypass data of References

26 and 27. Correla tions of Sun (Ref. 28) and Jones (Ref. 29) compare well to counter-current flow limiting (CCFL) test data of interest to BWRs. 3.4.2.2 Experimental Data for ECC Bypass. The following tests should be considered in establishing a set of data for scaled-down PWR downcomers:

"* Battelle Columbus test (Ref. 26) "* Creare test (Refs. 26 and 27) For a full-scale PWR vessel, ECC bypass data will become available from the forthcoming upper plenum test facility (UPTF) experiments performed as part of the 2D/3D program sponsored by the 1.157-7 Federal Republic of Germany, Japan, and the United States. For BWRs, the following test should be consid ered in establishing an acceptable set of relevant data: 0 SSTF test data (Refs. 30 through 32) 3.5 Noding Near the Break and ECCS Injection Point The break location and ECCS injection point are areas of high fluid velocity and complex fluid flow and contain phenomena that are often difficult to cal culate. The results of these calculations are often highly dependent on the noding. Sufficient sensitivity studies should be performed on the noding and other important parameters to ensure that the calculations provide realistic results.

3.6 Frictional Pressure Drop The frictional losses in pipes and other compo nents should be calculated using models that include variation of friction factor with Reynolds number and account for two-phase flow effects on friction.

Best estimate models will be considered acceptable pro vided their technical basis is demonstrated with ap propriate data and analyses.

3.6.1 Model Evaluation Procedure for Frictional Pressure Drop A model for frictional pressure drop to be used in ECCS evaluation should: a. Be checked against a set of relevant data, and b. Be consistent with models used for calculat ing gravitational and acceleration pressure drops. If void fraction models or correlations used to calculate the three components of the total pressure drop differ one from another, a quantitative justification must be provided.

Uncertainties and bias of a correlation or model should be stated as well as the range of applicability.

3.6.2 Experimental Data for Frictional Pressure Drop An acceptable set of relevant data should cover, as far as possible, the ranges of parameters (mass flux, quality, pressure, fluid physical properties, roughness, and geometries)

that are found in actual plant applications.

The following tests should be con sidered in establishing an acceptable set of relevant data:

• Vertical tubes CISE test (Refs. 33 and 34)* Horizontal tubes GE tests (Refs. 35 and 36)

• Rod bundles GE tests (Ref. 37) 3.7 Momentum Equation The following effects should be taken into account in the two-phase conservation of momentum equation:

(1) temporal change in momentum, (2) momentum convection, (3) area change momen tum flux, (4) momentum change due to compressibil ity, (5) pressure loss resulting from wall friction, (6) pressure loss resulting from area change, and (7) gravitational acceleration.

Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.8 Critical Heat Flux Best-estimate models developed from appropri ate steady-state or transient experimental data should be used in calculating critical heat flux (CHF) during loss-of-coolant accidents.

The codes in which these models are used should contain suitable checks to en sure that the range of conditions over which these correlations are used are within those intended.

Re search has shown that CHF is highly dependent on the fuel rod geometry, local heat flux, and fluid con ditions. After CHF is predicted at an axial fuel rod location, the calculation may use nucleate boiling heat transfer correlations if the calculated local fluid and surface conditions justify the reestablishment of nucleate boiling. Best-estimate models will be consid ered acceptable provided their technical basis is dem onstrated with appropriate data and analyses.

3.9 Post-CHF Blowdown Heat Transfer Models of heat transfer from the fuel to the sur rounding fluid in the post-CHF regimes of transition and film boiling should be best-estimate models based on comparison to applicable steady-state or transient data. Any model should be evaluated to demonstrate that it provides acceptable results over the applicable ranges. Best-estimate models will be considered ac ceptable provided their technical basis is demon strated with appropriate data and analyses.

3.9.1 Model Evaluation Procedure for Post-CHF Heat Transfer A model to be used in ECCS evaluation to calcu late post-CHF heat transfer from rod bundles should: a. Be checked against an acceptable set of rele vant data, and b. Recognize effects of liquid entrainment, ther mal radiation, thermal nonequilibrium, low and high mass flow rates, low and high power 1.157-8 densities, and saturated and subcooled inlet conditions.

The uncertainties and bias of models or correla tions used to calculate post-CHF heat transfer should be stated as well as the range of their applicability.

3.9.2 Experimental Data for Post-CHF Heat Transfer The acceptable set of relevant data should cover power densities, mass flow rates, fluid conditions, and rod bundle geometries pertinent to light-water re actor designs and applications.

The following tests should be considered in establishing an acceptable set of relevant data: "* ORNL tests (Refs. 38 and 39) "* FLECHT-SEASET

tests (Ref. 40) "* INEL tests (Ref. 41) "* ORNL data base (Ref. 42) 3.9.3 Post-CHF Heat Transfer from Uncovered Bundles During some time periods of small-break loss-of coolant accidents and during portions of large breaks prior to reflood, partial or complete core uncovering may be calculated to occur. Under these circum stances, special considerations for calculating heat transfer are necessary.

3.9.3.1 Model Evaluation Procedures for Heat Transfer from Uncovered Rod Bundles. A correlation to be used in ECCS evaluations to calculate heat transfer from uncovered rod bundles should: a. Be checked against an acceptable set of rele vant data, and b. Recognize the effects of radiation and of laminar, transition, and turbulent flows. Uncertainties and bias in the models and correla tions used to calculate post-CHF heat transfer should be stated, as should the range of their applicability.

The correlation derived should include a stated procedure for correcting for radiative heat transfer and for estimating the vapor temperatures.

The Hot tel procedure cited in Reference

43 is a satisfactory example.

The turbulent correlation may be of the general form: Nu = A Rem Prn for higher Reynolds numbers (Re), where the coeffi cients A, m, and n are modifications from the basic Dittus-Boelter form and may be functions of other variables.

Pr represents the Prandtl number, and Nu is the Nusselt number. The physical properties may be defined as wall, film, or vapor values. A distinction from, and transition to, laminar convection (i.e., Re <2000) should be made, with a value of the laminar heat transfer for rod bundles that is appropriate for the applicable bundle geometry and flow conditions.

Other forms and values, depending on the bun dle geometry and flow conditions, are also appropri ate. 3.9.3.2 Experimental Data for Heat Transfer from Uncovered Rod Bundles. An acceptable set of relevant data for post-CHF heat transfer from uncovered rod bundles should cover power densities, fluid conditions, and rod bundle geometries pertinent to light-water reactor design and application.

The following tests should be considered in establishing an acceptable set of relevant data: "* ORNL-THTF

tests (Refs. 43 and 44) "* FLECHT-SEASET

tests (Refs. 45 and 46) "* ORNL data base (Ref. 42) 3.10 Pump Modeling The characteristics of rotating primary system pumps should be derived from a best-estimate dy namic model that includes momentum transfer be tween the fluid and the rotating member, with vari able pump speed as a function of time. The pump model resistance and other empirical terms should be justified through comparisons with applicable data. The pump model for the two-phase region should be verified by comparison to applicable two-phase per formance data. Pump coastdown following loss of power should be treated in a best-estimate manner. A locked rotor following a large-break loss-of-coolant accident need not be assumed unless it is calculated to occur. Best-estimate models will be considered ac ceptable provided their technical basis is demon strated with appropriate data and analyses.

3.11 Core Flow Distribution During Blowdown The core flow through the hottest region of the core during the blowdown should be calculated as a function of time. For the purpose of these calcula tions, the hottest region of the core should not be greater than the size of one fuel assembly.

Calcula tions of the flow in the hot region should take into account any cross-flow between regions and any flow blockage calculated to occur during the blowdown as a result of cladding swelling or rupture. The numeri cal scheme should ensure that unrealistic oscillations of the calculated flow do not result. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.1.157-9

3.12 Post-Blowdown Phenomena 3.12.1 Containment Pressure The containment pressure used for evaluating cooling effectiveness during the post-blowdown phase of a loss-of-coolant accident should be calculated in a best-estimate manner and should include the effects of containment heat sinks. The calculation should in clude the effects of operation of all pressure-reducing equipment assumed to be available.

Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.12.2 Calculation of Post-Blowdown Thermal Hydraulics for Pressurized Water Reactors The refilling of the reactor vessel and the ulti mate reflooding of the core should be calculated by a best-estimate model that takes into consideration the thermal and hydraulic characteristics of the core, the emergency core cooling systems, and the primary and secondary reactor systems. The model should be ca pable of calculating the two-phase level in the reactor during the postulated transient.

Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.12.2.1 Model Evaluation Procedures for Post-Blowdown Thermal Hydraulics.

A correlation or model to be used in ECCS evaluation to calculate level swell should be checked against an acceptable set of relevant data and should recognize the effects of depressurization, boil-off, power level, fluid conditions, and system geometry.

The correlation proposed by Chexal, Horowitz, and Lellouche (Ref. 47) provides acceptable results when compared to experimental data reported in References

43, 48, 49, and 50. Uncertainties and bias of a correlation or model used to calculate level swell should be stated, as should the range of applicability.

The primary coolant pumps should be assumed to be operating in the expected manner, based on the assumptions of Regulatory Position 3. 1, when calcu lating the resistance offered by the pumps to fluid flow. Models will be considered acceptable provided their technical basis is demonstrated through com parison with appropriate data and analyses.

The total fluid flow leaving the core exit (car ryover) should be calculated using a best-estimate model that includes the effect of cross-flow on car ryover and core fluid distribution.

Thermal-hydraulic phenomena associated with unique emergency core cooling systems, such as upper plenum injection and upper head injection, should be accounted for. The effects of the compressed gas in the accumulator fol lowing accumulator water discharge should be in cluded in the calculation.

Any model or code used for this calculation should be assessed against applica ble experimental dat

a. Reference

7 describes a large body of refill/reflood thermal-hydraulic data obtained from the 2D/3D program that is appropriate for consideration.

3.12.2.2 Experimental Data for Post Blowdown Thermal Hydraulics.

The following tests should be considered when establishing an acceptable set of relevant data: "* GE tests (Refs. 48 and 51) "* ORNL tests (Refs. 43 and 49) "* FLECHT-SEASET

test (Ref. 45) "* THETIS tests (Ref. 50) 3.12.3 Steam Interaction with Emergency Core Cooling Water in Pressurized Water Reactors The thermal-hydraulic interaction between the steam or two-phase fluid and the emergency core cooling water should be taken into account in calcu lating the core thermal hydraulics and the steam flow through the reactor coolant pipes during the time the accumulators are discharging water. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

3.12.4 Post-Blowdown Heat Transfer for Pressurized Water Reactors During refilling of the reactor vessel and ultimate reflooding of the core, the heat transfer calculations should be based on a best-estimate calculation of the fluid flow through the core, accounting for unique emergency core cooling systems. The calculations should also include the effects of any flow blockage calculated to occur as a result of cladding swelling or rupture. Heat transfer calculations that account for two-phase conditions in the core during refilling of the reactor vessel should be justified through com parisons with experimental data. Best-estimate mod els will be considered acceptable provided their tech nical basis is demonstrated through comparison with appropriate data and analyses.

The FLECHT-SEASET

tests (Refs. 40, 45, and 46) should be considered when establishing an ac ceptable set of relevant dat

a. Reference

7 contains extensive information regarding a large amount of ex perimental reflood heat transfer data. This informa tion should also be considered when developing and assessing models. The results from the 2D/3D pro gram are particularly relevant.1.157-10

3.13 Convective Heat Transfer Coefficients for Boiling Water Reactor Rods Under Spray Cooling Models will be considered acceptable provided their technical bases can be justified with appropriate

-data and analyses.

These models should contain the following:

1. Following the blowdown period, convective heat transfer coefficients should be deter mined based on the calculated fluid condi tions and heat transfer modes within the bun dle and on the calculated rod temperatures.

2. During the period following the flashing of the lower plenum fluid, but prior to ECCS initiation, heat transfer models should in clude cooling by steam flow or by a two phase mixture, if calculated to occur. 3. Following initiation of ECCS flow, but prior to reflooding, heat transfer should be based on the actual calculated bundle fluid condi tions and best-estimate heat transfer models that take into account rod-to-rod variations in heat transfer.

4. After the two-phase reflood level reaches the level under consideration, a best-estimate heat transfer model should be used. This model should include the effects of any flow blockage calculated to occur as a result of cladding swelling or rupture.

5. Thermal-hydraulic models that do not calcu late multiple channel effects should be com pared with applicable experimental data or more detailed calculations to ensure that all important phenomena are adequately calculated.

3.14 Boiling Water Reactor Channel Box Under Spray Cooling Following the blowdown period, heat transfer from the channel box and wetting of the channel box should be based on the calculated fluid conditions on both sides of the channel box and should make use of best-estimate heat transfer and rewetting models that have been compared with applicable experimen tal data. 3.15 Special Considerations for a Small-Break Loss-of-Coolant Accident in Pressurized Water Reactors The slower small-break loss-of-coolant accident leads to fluid conditions characterized by separation of the fluid phases versus the more homogeneous fluid conditions that would result from rapid large break loss-of-coolant accident transients.

Phenomena

~ that would occur in a PWR during a small-break loss of-coolant accident would, therefore, be significantly different from those phenomena that would occur during a large-break loss-of-coolant accident.

The distribution of liquid throughout the reactor system, in addition to the total liquid inventory, is of in creased importance for the small-break loss-of coolant accident.

A number of special factors must be given increased consideration in small-break loss of-coolant accident calculations to correctly predict phenomena influenced by the liquid inventory distribution.

Break flow may be greatly influenced by the loca tion and specific geometry of the break. For a break in a horizontal pipe containing stratified flow, the quality of the break flow will be a strong function of the assumed location of the break on the pipe (e.g., top or bottom). Small-break loss-of-coolant accident calculations should, therefore, include various as sumed break locations in the spectrum of breaks ana lyzed. The assumed operating state of the reactor coolant pump will also influence the distribution of liquid throughout the system and the amount of liquid lost through the break. The pump operation assumptions used in the cal culations should be the most likely, based on operat ing procedures, with appropriate consideration of the uncertainty of the pump operation during an actual event. Level depression in the core region and subse quent core heatup may be influenced by liquid holdup in the steam generator tubes, manometric ef fects of liquid in the piping and loop seal region, and liquid levels relative to vent paths for steam through upper plenum bypass flow paths and vent valves. Steam generator heat transfer under "reflux" or "boiler-condensor" modes of operation may also strongly influence core inventory through level de pression and the effect on total system pressure and, thus, on ECCS flow. These phenomena should be carefully considered in the calculation.

Sensitivity studies of the importance of these effects should be performed for use in the uncertainty evaluation.

Heat transfer from an uncovered core under high-pressure conditions typical during a small-break loss-of-coolant accident may include contributions from both convective and radiation heat transfer to the steam. Models will be considered acceptable pro vided their technical basis is demonstrated through comparison with appropriate data and analyses.

Spe cific guidance regarding uncovered bundle heat trans fer is given in Regulatory Position 3.9.3. 3.16 Other Features of Best-Estimate Codes No list of best-estimate code features could be all-inclusive, because the important features of a best-estimate code may vary depending on the tran sient to be calculated and the required accuracy of the calculation.

Because of this, no attempt has been made to construct an exhaustive list of best-estimate code features.

Rather, features that were identified as 1.157-11 important for inclusion in Appendix K were used as a basis for the above list. These features are not neces sarily any more or less important than other code fea tures, but were highlighted because it is necessary to give specific examples of how current best-estimate models may vary from methods used traditionally in evaluation model codes using the various Appendix K conservatisms.

In addition, models have not been included for areas in which the best model would be highly dependent on the specific plant design or the specific transient under consideration.

The NRC staff believes that good examples of best-estimate thermal-hydraulic transient codes are those developed by the NRC (e.g., TRAC-PWR, TRAC-BWR, RELAPS, COBRA, and FRAP). Al though these codes are subject to further improve ment, based on their ongoing use and assessment, they currently provide reasonable best-estimate cal culations of the loss-of-coolant accident in a full-scale light-water reactor. This is substantiated through the code development and assessment literature gener ated by the NRC and its contractors over the past several years. It is possible, however, to describe in general how other features of best-estimate codes should be constructed.

Two basic criteria should be applied, completeness and comparisons to experimental data. 3.16.1 Completeness Best-estimate codes should contain models in sufficient detail to predict phenomena that are impor tant to demonstrate compliance with the acceptance criteria specified in paragraph

50.46(b) of 10 CFR Part 50 (e.g., peak cladding temperature).

Simplifi cations are acceptable as lopg as code uncertainties or biases do not become so large that they cast doubt on the actual behavior that would occur or on the true effect of assumed initial and boundary condi tions (e.g., equipment sizing, safety system settings).

Comparisons of the overall calculations to integral ex periments should be performed to ensure that impor tant phenomena can be predicted and to help in making judgments on the effect of code simplifica tions. Consideration should also be given to the un certainty and validity of the experiment to ensure that meaningful comparisons are being made. 3.16.2 Data Comparisons Individual best-estimate models should be com pared to applicable experimental data to ensure that realistic behavior is predicted and that relevant ex perimental variables are included.

Uncertainty analy ses are required to ensure that a major bias does not exist in the models and that the model uncertainty is small enough to provide a realistic estimate of the ef fect of important experimental variables.

Uncertainty analyses should also consider experimental uncer-tainty to ensure that meaningful comparisons are be ing made.

4. ESTIMATION

OF OVERALL CALCULATIONAL

UNCERTAINTY

4.1 General The term "uncertainty," when applied to best estimate thermal-hydraulic transient codes, is used at two levels. At the lower or more detailed level, the term refers to the degree to which an individual model, correlation, or method used within the code represents the physical phenomenon it addresses.

These individual uncertainties, when taken together, comprise the "code uncertainty." The combined uncertainty associated with indi vidual models (i.e., code uncertainty)

within the best estimate codes does not account for all of the uncer tainty associated with the model's use. In addition to the code uncertainty, various other sources of uncer tainty are introduced when attempting to use best estimate codes to predict full-scale plant thermal-hydraulic response.

These sources include uncertainty associated with the experimental data used in the code assessment process (including appli cability of the data to full-scale reactors), the input boundary and initial conditions, and the fuel behav ior. Additional sources of uncertainty stem from the use of simplifying assumptions and approximations.

A careful statement of these assumptions and approxi mations should be made, and the uncertainty associ ated with them should be taken into account. There fore, the "overall calculational uncertainty" is defined as the uncertainty arrived at when all the contributions from the sources identified above, in cluding the code uncertainty, are taken into account.

A 95% probability level is considered acceptable to the NRC staff for comparison of best-estimate pre dictions to the applicable limits of paragraph

50.46(b) of 10 CFR Part 50 to meet the requirement of paragraph

50.46(a)(1)(i)

to show that there is a high probability that the criteria will not be exceeded.

The basis for selecting the 95% probability level is primarily for consistency with standard engineering practice in regulatory matters involving thermal hy draulics.

Many parameters, most notably the depar ture from nucleate boiling ratio (DNBR), have been found acceptable by the NRC staff in the past at the 95% probability level. This 95% probability level would also be applied to small-break loss-of-coolant accidents, which have a higher probability than large breaks. The dominant factors influencing risk from small-break loss-of coolant accidents include equipment availability and operator actions. Calculational uncertainties are much less important than factors such as operator recognition of the event, the availability of equip ment, and the correct use of this equipment..

The use 1.157-12 of a best-estimate calculation with reasonable and quantifiable uncertainty is expected to provide a re duction in the overall risk from a small-break loss-of coolant accident by providing more realistic calcula tions with which to evaluate operator guidelines and -' determine the true effect of equipment availability.

Regulatory Position 3 provides a description of the features that should be included in the overall code uncertainty evaluation that is called for in para graph 50.46(a) (1). This uncertainty evaluation should make use of probabilistic and statistical meth ods to determine the code uncertainty.

For a calcula tion of this complexity, a completely rigorous mathe matical treatment is neither practical nor required.

In many cases, approximations and assumptions may be made to make the overall calculational uncertainty evaluation possible.

A careful statement of these as sumptions and approximations should be made so that the NRC staff may make a judgment as to the validity of the uncertainty evaluation.

The purpose of the uncertainty evaluation is to provide assurance that for postulated loss-of-coolant accidents a given plant will not, with a probability of 95% or more, ex ceed the applicable limits specified in paragraph

50.46(b).

4.2 Code Uncertainty This regulatory guide makes a distinction be tween the terms "code uncertainty" and "overall cal culational uncertainty." The latter term is defined above and includes the contributions to the uncer tainty described in Regulatory Positions

4.2 and 4.3. The components of the code uncertainty (i.e., the contribution to the overall uncertainty from the mod els and numerical methods used) are described in this section.

Code uncertainty should be evaluated through di rect data comparison with relevant integral systems and separate-effects experiments at different scales. In this manner, an estimate of the uncertainty attrib utable to the combined effect of the models and cor relations within the code can be obtained for all scales and for different phenomena.

Comparison to a sufficient number of integral systems experiments, from different test facilities and different scales, should be made to ensure that a reasonable estimate of code uncertainty and bias has been obtained.

When necessary, separate-effects experiments should be used to establish code uncertainty for specific phe nomena (e.g., comparisons to Cylindrical Core Test Facility data to ascertain code uncertainty in model ing upper plenum injection performance).

Code com parisons should account for limitations of the meas urements and calibration errors. These comparisons should be performed for im portant key parameters to demonstrate the overall best-estimate capability of the code. For large-break loss-of-coolant accidents, the most important key pa rameter is peak cladding temperature, which is ad dressed by one of the criteria of paragraph

50.46(b) and has a direct influence on the other criteria.

In addition, a code uncertainty evaluation should be performed for other important parameters for the transient of interest to evaluate compensating errors. For small-break loss-of-coolant accidents, the clad ding temperature response is the most important parameter;

however, the ability of the codes to pre dict overall system mass and reactor vessel inventory distribution should also be statistically examined.

In evaluating the code uncertainty, it will be nec essary to evaluate the code's predictive ability over several time intervals, since different processes and phenomena occur at different intervals.

For example, in large-break loss-of-coolant accident evaluations, separate code uncertainties may be required for the peak cladding temperature during the blowdown and post-blowdown phases. Justification for treating these uncertainties individually or methods for combining them should be provided.

The experimental information used to determine code uncertainty will usually be obtained from facili ties that are much smaller than nuclear power reac tors. Applicability of these results should be justified for larger scales. The effects of scale can be assessed through comparisons to available large-scale separate effects tests and through comparison to integral tests from various sized facilities.

If there are scaling prob lems, particularly if predictions are nonconservative, the code should be improved for large-scale plants (i.e., nuclear reactors).

Codes not having scaling ca pability will not be acceptable if their predictions are nonconservative.

4.3 Other Sources of Uncertainty When a best-estimate methodology is used to predict reactor transients, sources of uncertainty other than the limitations in the individual models and numerical methods (i.e., code uncertainty)

are introduced.

The following contributors to the overall calculational uncertainty should also be considered in the uncertainty analysis.

4.3.1 Initial and Boundary Conditions and Equipment Availability When a plant input model is prepared, certain relationships describing the plant boundary and initial conditions and the availability and performance of equipment are defined. These include factors such as initial power level, pump performance, valve activa tion times, and control systems functioning.

Uncer tainties associated with the boundary and initial con ditions and the characterization and performance of equipment should be accounted for in the uncertainty evaluation.

It is also acceptable to limit the variables 1.157-13 to be considered by setting their values to conserva tive bounds. 4.3.2 Fuel Behavior Variability of the results of plant transient calcu lations can result from uncertainties associated with fuel behavior, which are not included in the compari sons of code results with integral experiments since most integral tests use electrically heated rods. This uncertainty includes many effects such as fuel con ductivity, gap width, gap conductivity, and peaking factors. These uncertainties should be quantified and used in the determination of the overall calculational uncertainty.

4.3.3 Other Variables There may be individual models within the best estimate code whose effect may not have been evalu ated by the comparison to the integral systems data. For example, since most integral systems experiments use electrically heated rods, uncertainties associated with the prediction of core decay heat and cladding metal/water reaction have not been evaluated.

In ad dition, to demonstrate the overall adequacy of the predictive ability of the best-estimate code, it may be necessary to use empirically arrived at break discharge coefficients to obtain a reasonable break flow. The uncertainties in the individual models that have not been evaluated by comparison to integral systems data should be quantified and used in the determination of overall code uncertainty.

4.4 Statistical Treatment of Overall Calculational Uncertainty The methodology used to obtain an estimate of the overall calculational uncertainty at the 95% prob ability limit should be provided and justified.

If linear independence is assumed, suitable justification should be provided.

The influence of the individual parameters on code uncertainty should be examined by making comparisons to relevant experimental data. Justification should be provided for the as sumed distribution of the parameter and the range considered.

In reality, the true statistical distribution for the key parameters (e.g., peak cladding temperature)

is unknown. The choice of a statistical distribution should be verified using applicable engineering data and information.

The statistical parameters appropri ate for that distribution should be estimated using available data and results of engineering analyses.

Supporting documentation should be provided for this selection process. These estimated values are as sumed to be the true values of the statistical param eters of the distribution.

With these assumptions, an upper one-sided probability limit can be calculated at the 95% level. As the probability limit approaches

2200'F, more care must be taken in the selection and justification of the statistical distribution and in the estimation of its statistical parameters.

If a normal distribution is selected and justified, the probability limit can be conservatively calculated using two standard deviations.

The added conservatism of the two standard deviations compared to the 95th percentile is used to account for uncertainty in the probability distribution.

Other techniques that account for the uncertainty in a more detailed manner may be used. These techniques may require the use of confidence levels, which are not required by the above approach.

The evaluation of the peak cladding temperature at the 95% probability level need only be performed for the worst-case break identified by the break spec trum analysis in order to demonstrate conformance with paragraph

50.46(b).

However, in order to use this approach, justification must be provided that demonstrates that the overall calculational uncer tainty for the worst case bounds the uncertainty for other breaks within the spectrum.

It may be neces sary to perform separate uncertainty evaluations for large- and small-break loss-of-coolant accidents be cause of the substantial difference in system thermal hydraulic behavior.

The revised paragraph

50.46(a) (1)(i) requires that it be shown with a high probability that none of the criteria of paragraph

50.46(b) will be exceeded, and is not limited to the peak cladding temperature criterion.

However, since the other criteria are strongly dependent on peak. cladding temperature, explicit consideration of the probability of exceeding the other criteria may not be required if it can be demonstrated that meeting the temperature criterion at the 95% probability level ensures with an equal or greater probability that the other criteria will not be exceeded.

4.5 NRC Approach to LOCA Uncertainty Evaluation Chapter 4 of the "Compendium of ECCS Research for Realistic LOCA Analysis" (Ref. 7) presents a methodology that has been used for evaluating the overall calculational uncertainty in peak cladding temperature predictions for best-estimate thermal-hydraulic transient codes that the NRC has developed.

D. IMPLEMENTATION

The purpose of this section is to provide informa tion to applicants and licensees regarding the NRC staff's plans for using this regulatory guide. Licensees and applicants may propose means other than those specified by the provisions of the Regulatory Position of this guide for meeting applica ble regulations.

This guide has been approved for use by the NRC staff as an acceptable means of 1.157-14 complying with the Commission's regulations and for evaluating submittals in the following categories:

1. Construction permit applicants that choose to make use of the provisions of § 50.46 that allow the use of realistic models as an alter native to the features of Appendix K of 10 CFR Part 50. 2. Operating license applicants that choose to make use of the provisions of § 50.46 that allow the use of realistic models as an alter native to the features of Appendix K of 10 CFR Part 50. 3. Operating reactor licensees will not be evaluated against the provisions of this guide except for new submittals that make use of the provisions of § 50.46 that allow the use of realistic models as an alternative to the features of Appendix K of 10 CFR Part 50.1.157-15 REFERENCES

1. Los Alamos National Laboratory, "TRAC PF1/MOD1:

An Advanced Best-Estimate Com puter Program for Pressurized Water Reactor Thermal-Hydraulic Analysis," NUREG/ CR-3858 (LA-10157-MS), July 1986. 2. Idaho National Engineering Laboratory, "TRAC-BD1/MOD1:

An Advanced Best Esti mate Computer Program for Boiling Water Re actor Transient Analysis," NUREG/CR-3633, 4 Vols. (EGG-2294), April 1984. 3. Idaho National Engineering Laboratory, "RELAP5/MOD2 Code Manual," Vols. 1 & 2, NUREG/CR-4312, August 1985. (Available in the NRC Public Document Room.) 4. Pacific Northwest Laboratory, "COBRA/TRAC

-A Thermal-Hydraulics Code for Transient Analysis of Nuclear Reactor Vessels and Pri mary Coolant Systems," NUREG/CR-3046, 5 Vols. (PNL-4385), March 1983. 5. L. J. Siefken et al., "FRAP-T6:

A Computer Code for the Transient Analysis of Oxide Fuel Rods," NUREG/CR-2148 (EG&G, EGG-2104), May 1981. 6. G. A. Berna et al., "FRAPCON-2:

A Com puter Code for the Calculation of Steady State Thermal-Mechanical Behavior of Oxide Fuel Rods," NUREG/CR-1845, January 1981. 7. "Compendium of ECCS Research for Realistic LOCA Analysis," NUREG-1230, December 1988. 8. D. Lanning and M. Cunningham, "Trends in Thermal Calculations for Light Water Reactor Fuel (1971-1981)," in Ninth Water Reactor Safety Research Information Meeting, USNRC, NUREG/CP-0024, Vol. 3, March 1982. 9. Idaho National Engineering Laboratory, "MATPRO Version 11 (Revision

2): A Hand book of Materials Properties for Use in the Analysis of Light-Water Reactor Fuel Rod Be havior," NUREG/CR-0497, Rev. 2, August 1981. 10. American Nuclear Society, "American Na tional Standard for Decay Heat Power in Light Water Reactors," ANSI/ANS-5.1-1979, August 1979. (ANS, 555 North Kensington Avenue, La Grange Park, Illinois 60525.) 11. J. V. Cathcart et al., "Zirconium Metal-Water Oxidation Kinetics:

IV Reaction Rate Studies," Oak Ridge National Laboratory, ORNL/NUREG-17, August 1977. (Available from NTIS.) 12. H. J. Richter, "Separated Two-Phase Flow Model: Application to Critical Two-Phase Flow," EPRI Report NP-1800, Electric Power Research Institute, Palo Alto, CA, April 1981. 13. D. Abdollahian et al., "Critical Flow Data Re view and Analysis," Report NP-2192, Electric Power Research Institute, Palo Alto, CA, Janu ary 1982. 14. USNRC, "The Marviken Full Scale Critical Flow Tests, Summary Report," (Joint Reactor Safety Experiments in the Marviken Power Sta tion, Sweden), NUREG/CR-2671, May 1982. 15. M. Reocreux, "Contribution to the Study of Two-Phase Steam-Water Critical Flow," Ph.D. Thesis, L'Universite Scientifique Medicale de Grenoble, 1974. (English translation available from NTIS, LIB/Trarns-576.)

16. N. Abuaf, G. A. Zimmer, B. J. C. Wu, "A Study of Nonequilibrium Flashing of Water in a Converging-Diverging Nozzle," NUREG/ CR-1864, Vols. 1-2 (Brookhaven National Laboratory, BNL-NUREG-51317), March 1982. 17. G. L. Sozzi and W. A. Sutherland, "Critical Flow of Saturated and Subcooled Water at High Pressure," General Electric Company, GE Re port NEDO-13418, 1975. (Available in the NRC Public Document Room.) 18. R. A. Edwards and T. P. O'Brien, "Studies of Phenomena Connected with the Depressuriza tion of Water Reactois," Nuclear Energy (Jour nal of the British Nuclear Energy Society), Vol. 9, No. 2, April 1970. 19. Commissariat a L'Energie Atomique, C. Jean dey et al., "Auto vaporisation d'ecoulements eau/vapeur," Report TT, No. 163, Centre d'Etudes Nucleaires de Grenoble, Dept. des Reacteurs a Eau, Service des Transferts Ther miques, Grenoble, France, July 1981. (Copies may be obtained from Maurice Gomolinski, CEA, B.P. No. 6, 92260 Fontenay-aux-Roses Cedex, France.) 20. C. Jeandey and L. Gros d'Aillon, "Critical Flows in a Short Super Moby Dick Pipe," Rap port TT/SETRE/7

1, Centre d'Etudes Nucleaires de Grenoble, Grenoble, France, September

198

3. NRC Translation

1401 available from the NRC Public Document Room (52 FR 6334), ac cession number 8704060298.

1.157-16

21. J. L. Anderson and W. A. Owca, "Data Report for the TPFL Tee/Critical Flow Experiments," NUREG/CR-4164 (EG&G Idaho, Inc., EGG-2377), November 1985. S 22. J. Reimann and M. Khan, "Flow Through a Small Break at the Bottom of a Large Pipe with Stratified Flow," Nuclear Science and Engineer ing, Vol. 88, pp. 297-310, November 1984. 23. V. E. Schrock et al., "Steam-Water Critical Flow Through Small Pipes from Stratified Up stream Regions," in Heat Transfer 1986; C. L. Tien, V. P. Carey, and J. K. Ferrell, Editors; Vol. 5, pp. 2307-2311;

Hemisphere Publishing Corp., 242 Cherry St., Philadelphia, PA 19106, 1986. 24. V. E. Schrock et al., "Small Break Critical Dis charge -Roles of Vapor and Liquid Entrain ment in Stratified Two-Phase Region Upstream of the Break," NUREG/CR-4761 (Lawrence Berkeley Laboratory, LBL-22024), December 1986. 25. W. D. Beckner and J. N. Reyes, Research In formation Letter No. 128, "PWR Lower Plenum Refill Research Results," USNRC, December 8, 1981. (Available in the NRC Public Document Room.) 26. W. D. Beckner, J. N. Reyes, R. Anderson, "Analysis of ECC Bypass Data," U.S. Nuclear Regulatory Commission, NUREG-0573, July 1979. 27. C. J. Crowley et al., "1/5-Scale Countercurrent Flow Data Presentation and Discussion," NUREG/CR-2106 (Creare Incorporated, Creare TN-333), November 1981. 28. K. H. Sun, "Flooding Correlations for BWR Bundle Upper Tieplate and Bottom Side-Entry Orifices," in Multi-Phase Transport:

Funda mentals, Reactor Safety, Applications, Vol. 1, T. N. Veziroglu, Editor, Hemisphere Publishing Corp., 242 Cherry St., Philadelphia, PA 19106, 1979. 29. D. D. Jones, "Subcooled Counter-Current Flow Limiting Characteristics of the Upper Region of a BWR Fuel Bundle," General Electric Com pany, NEDG-NUREG-23549, July 1977. (Available in the NRC Public Document Room.)30. J. A. Findlay, "BWR Refill-Reflood Program Task 4.4 -CCFL/Refill System Effects Tests (30' Sector). Evaluation of Parallel Channel Phenomena," NUREG/CR-2566 (General Elec tric Company, GEAP-22044, EPRI NP-2373), November 1982.31. D. G. Schumacher et al., "BWR Refill-Reflood Program Task 4.4 -CCFL/Refill System Ef fects Tests (300 Sector). SSTF Systems Re sponse Test Results," NUREG/CR-2568 (Gen eral Electric Company, GEAP-22046, EPRI NP-2374), April 1983. 32. J. A. Findlay, "BWR Refill-Reflood Program Task 4.4 -CCFL/Refill System Effects Tests (300 Sector). Evaluation of ECCS Mixing Phe nomena," NUREG/CR-2786 (General Electric Company, GEAP-22150, EPRI NP-2542), May 1983. 33. G. P. Gaspari, C. Lombardi, G. Peterlongo, "Pressure Drops in Steam-Water Mixtures.

Round Tubes Vertical Upflow," Centro Infor mazioni Studi Esperienze, Milan, Italy, CISE-R83, 1964. (Available from NTIS.) 34. A. Alessadrini, G. Peterlongo, R. Ravetta, "Large Scale Experiments on Heat Transfer and Hydrodynamics with Steam-Water Mixtures.

Critical Heat Flux and Pressure Drop Measure ments in Round Vertical Tubes at the Pressure of 51 kg/cm 2 abs," Centro Informazioni Studi Esperienze, Milan, Italy, CISE-R86, 1963. (Available from NTIS.) 35. E. Janssen and J. A. Kervinen, "Two-Phase Pressure Drop Across Contractions and Expan sions: Water-Steam Mixtures at 600 to 1400 psia," AEC R&D Report GEAP-4622, 1964. (Available in the NRC Public Document Room.) 36. E. Janssen and J. A. Kervinen, "Two-Phase Pressure Drop in Straight Pipes and Channels:

Water-Steam Mixtures at 600 to 1400 psia," AEC R&D Report GEAP-4616, 1964. (Avail able in the NRC Public Document Room.) 37. R. T. Lahey, B. S. Shiralkar, D. W. Radcliffe, "Two-Phase Flow and Heat Transfer in Multi rod Geometries:

Subchannel and Pressure Drop Measurements in a Nine-Rod Bundle for Diabatic and Adiabatic Conditions," AEC R&D Report GEAP-13049, General Electric Com pany, March 1970. 38. G. L. Yoder et al., "Dispersed Flow Film Boil ing in Rod Bundle Geometry -Steady State Heat Transfer Data and Correlation Compari sons," NUREG/CR-2435 (Oak Ridge National Laboratory, ORNL-5822), April 1982. 39. D. G. Morris et al., "Dispersed Flow Film Boil ing of High Pressure Water in a Rod Bundle," NUREG/CR-2183 (Oak Ridge National Labora tory, ORNL/TM-7864), September

1982. 40. N. Lee et al., "PWR FLECHT-SEASET

Un blocked Bundle, Forced and Gravity Reflood Task Data Evaluation and Analysis Report," 1.157-173/41ý~

NUREG/CR-2256 (Westinghouse Electric Cor poration, WCAP-9891, EPRI NP-2013), No vember 1981. 41. R. C. Gottula et al., "Forced Convective, Non equilibrium, Post-CHF Heat Transfer Experi ment Data and Correlation Comparison Re port," NUREG/CR-3193 (EG&G Idaho, Inc., EGG-2245), April 1985. 42. G. L. Yoder, "Rod Bundle Film Boiling and Steam Cooling Data Base and Correlation Evaluation," NUREG/CR-4394 (Oak Ridge Na tional Laboratory, ORNL/TM-9628), August 1986. 43. T. M. Anklam et al., "Experimental Investiga tions of Uncovered-Bundle Heat Transfer and Two-Phase Mixture-Level Swell Under High Pressure Low Heat-Flux Conditions," NUREG/ CR-2456 (Oak Ridge National Laboratory, ORNL-5848), April 1982. 44. G. L. Yoder et al., "High Dryout Quality Film Boiling and Steam Cooling Heat Transfer Data from a Rod Bundle," NUREG/CR-3502 (Oak Ridge National Laboratory, ORNL/TM-8794), January 1984. 45. S. Wong and L. E. Hochreiter, "Analysis of the FLECHT-SEASET

Unblocked Bundle Steam Cooling and Boiloff Tests," NUREG/CR-1533 (Westinghouse Electric Corporation, WCAP-9729, EPRI NP-1460), March 1981. 46. M. J. Loftus et al., "PWR FLECHT SEASET 21-Rod Bundle Flow Blockage Test Data and Analysis Report," NUREG/CR-2444, Vol. 1-2 (Westinghouse Electric Corporation, WCAP-9992, EPRI NP-2014), September

1982. 47. B. J. Chexal, J. Horowitz, G. Lellouche, "An Assessment of Eight Void Fraction Models for Vertical Flows," NSAC-107, Electric Power Re-search Institute, Palo Alto, CA, December 1986. 48. D. S. Seely and R. Muralidharan, "BWR Low Flow Bundle Uncovery Tests and Analysis," NUREG/CR-2231 (General Electric Company, GEAP-24964, EPRI NP-1781), April 1982. 49. T. M. Anklam, "ORNL Small-Break LOCA Heat Transfer Series I: Two-Phase Mixture Level Swell Results," NUREG/CR-2115 (Oak Ridge National Laboratory, ORNL/NUREG/

TM-447), September

1981. 50. D. Jowitt, "A New Voidage Correlation for Level Swell Conditions," Winfrith UK, AEEW-R-1488, December 1981. (Available in the NRC Public Document Room.) 51. J. A. Findlay, "BWR Refill-Reflood Program Task 4.8 -Model Qualification Task Plan," NUREG/CR-1899 (General Electric Company, GEAP-24898, EPRI NP-1527), August 198

1. ADDRESSES

NUREG- and NUREG/CR-series documents are available from the Government Printing Office (GPO) and the National Technical Information Service (NTIS). U.S. Government Printing Office Post Office Box 37082 Washington, DC 20013-7082 National Technical Information Service Springfield, VA 22161 Documents that are in the NRC Public Document Room are available for inspection or copying for a fee. USNRC Public Document Room 2120 L Street NW. Washington, DC 1.157-18 REGULATORY

ANALYSIS A separate regulatory analysis has not been pre> pared in support of this regulatory guide. The regula tory analysis that supports the rulemaking effort also covers this regulatory guide. "Regulatory Analysis for Revision of the ECCS Rule and Supporting- Regula tory Guide," is available in the NRC Public Docu ment Room, 2120 L Street NW., Washington, DC, under Regulatory Guide 1.157 (52 FR 6334).1.157-19 UNITED STATES NUCLEAR REGULATORY

COMMISSION

WASHINGTON, D.C. 20555 OFFICIAL BUSINESS PENALTY FOR PRIVATE USE, $300 FIRST CLASS MAIL POSTAGE &" FEES PAID USNRC PERMIT No. G-67