struction permit and operating license applications with regard to ECCS analyses.

ANALYTICAL ECCS REQUIREMENTS

2. Analytical requirements: A second major feature of the new acceptance criteria is its requirements for analyzing ECCS performance. New section 50.46 requires that system performance be calculated in accordance with an acceptable evaluation model-whose features are described in a new appendix K-and that the performance be calculated for a number of postulated loss-of-coolant accidents of different sizes, locations, and other properties sufficient to provide assurance that the entire spectrum of postulated loss-of-coolant accidents is covered. The largest rupture to be considered is a piping break equivalent in size to the double-ended rupture of the largest pipe in the reactor coolant system.

EVALUATION MODELS

3. Evaluation models: The third major subject covered in the new acceptance criteria is required and acceptable features of the evaluation models to be used in determining compliance with the specified ECCS performance criteria.

The major topics covered in new appendix K, wherein these features are specified, are sources of heat during the LOCA, swelling and rupture of the cladding and fuel rod thermal parameters, blowdown phenomena, and post-blowdown phenomena and heat removal by the ECCS.

Appendix K also requires that a description of each evaluation model be furnished to the AEC for review, and describes the required specificity and other features of the model related to documentation of its technical adequacy. Later portions of this testimony will cover this aspect of the new criteria in more detail.

APPLICATION OF CRITERIA TO VARIOUS CLASSES OF LWR'S

4. Applicability: The final major feature of the Acceptance Criteria is its provisions in new section 50.46 relating to its applicability to various classes of light water reactors, grouped in accordance with the time of issuance of construction permits or operating licenses. The Chairman has already discussed the implementation schedule specified in this new section.

ECCS PERFORMANCE CRITERIA

The new criteria specify five yardsticks against which the performance of the ECCS for all light water reactors are to be measured. In general, the new criteria are somewhat more restrictive and more specific than those of the interim acceptance criteria.

1. Peak cladding temperature and maximum oxidation criteria: The first two criteria specify that the calculated maximum fuel element cladding temperature during a postulated LOCA shall not exceed 2,200° F and that the calculated total oxidation of the cladding

shall not exceed 17 percent of the total cladding thickness before oxidation.

ZIRCALOY CLADDING

The purpose of these criteria is to assure that the Zircaloy cladding will remain sufficiently intact to retain the UO 2 fuel pellets in their separate fuel rods and therefore remain in an easily coolable array. Conservative calculations indicate that during some postulated LOCA's the cladding of some of the fuel rods would swell and burst locally with a longitudinal split. The split cladding would nevertheless remain in one piece if it were not too heavily oxidized, thus restraining the UO 2 pellets.

However, the possibility of further damage to the Zircaloy cladding later in the course of a LOCA must also be taken into account, since oxidation at high temperatures in the steam atmosphere could render the cladding brittle.

The limits specified in these criteria will assure that some ductility would remain in the Zircaloy cladding as it is quenched by water provided by the ECCS, and therefore that the core would remain essentially intact, in a geometry amenable to long-term_cooling.

The temperature limit of 2,200° represents a 100° decrease in the previously acceptable maximum calculated cladding temperature during a LOCA. The basis for this change is experimental information developed by Oak Ridge National Laboratory which indicated that the ductility of oxidized Zircaloy depends not only upon the temperature at which oxidation occurs, but also on the extent of oxidation, and that the relation between ductility and extent of oxidation appeared to change at temperatures above 2,200°.

While others have also observed that the resistance to Zircaloy cladding rupture depends both on the oxidation temperature and its extent, they do not share the belief that a 2,200° limit is necessary to assure adequate cladding ductility during a LOCA.

Nevertheless, until further experiments are conducted and analyzed, the Commission thought it prudent to reduce the maximum allowable calculated cladding temperature during a LOCA to a value of 2,200°.

There is widespread agreement that the total oxidation of the Zircaloy cladding during a LOCA should be limited to 17 percent as specified in the second new criterion. Such a limit was not expressly set forth in the previous criteria, but was an implicit limit for the reactor designs then under review.

Explicit specification of this maximum allowable value will avoid any future misunderstanding with regard to this important consideration affecting cladding ductility.

MAXIMUM HYDROGEN GENERATION CRITERION

This criterion requires that the calculated total amount of hydrogen generated from the chemical reaction of the cladding with water or steam be limited to 1 percent of the hypothetical amount that would

be generated if all the metal in the cladding cylinders surrounding the fuel, excluding the cladding surrounding the plenum volume, were to react.

The objective of this criterion is to assure that the hydrogen would not be generated during a LOCA in amounts that could lead to explosive concentrations in the reactor containment.

The criterion is essentially the same as in the previous interim acceptance criteria, but is more explicit in detailing how much of the Zircaloy is to be used as the basis for the 1-percent calculation.

COOLABLE GEOMETRY CRITERION

This criterion requires that calculated changes in core geometry be such that the core remains amenable to cooling. In view of the fundamental and historical importance of maintaining core coolability, this criterion was retained essentially unchanged from the interim acceptance criteria. However, as explained below, explicit analysis of geometry changes is required by the new evaluation models.

LONG-TERM COOLING CRITERION

This criterion requires that after calculated successful initial operation of the ECCS, the calculated core temperature be maintained at an acceptable low level and that decay heat be removed for the extended period of time required by the long-lived radioactivity remaining in the core. The purpose of this criterion is self-evident and noncontroversial.

REQUIRED FEATURES OF EVALUATION MODELS

The new regulation effects changes in two aspects of the 1971 interim acceptance criteria. These are: (a) the performance criteria discussed above; and (b) the evaluation models.

An evaluation model is a calculational framework for evaluating the behavior of the reactor system during a LOCA. In other words, an evaluation model contains a set of mathematical equations used for predicting the course and consequences of a LOCA and hence is a tool for judging the performance of an ECCS.

Each evaluation model contains several computer programs which solve the mathematical equations representative of the physical phenomena which occur during a LOCÂ; for example, flow of water; fission of atoms; heat transfer of convection, conduction, and radiation, et cetera.

These computer programs also contain mathematical representations of the reactor equipment; for example, fuel elements, pressure vessel, piping, instruments, valves, pumps, et cetera.

The solution of these equations simulates the response of a power reactor, starting from normal operation, through a LOCA, and up to the time of recovery from the accident. The evaluation model also includes values of important parameters used in the calculations.

COMPLEX MATHEMATICAL AND ENGINEERING METHODS REQUIRED

A large nuclear power reactor is a complex machine. Therefore, complex mathematical and engineering methods are required to calculate a reactor's response to a LOCA. For the most part, those complex methods are not subject to controversy or to large uncertainties because they are only applications of known contemporary technology to the nuclear field.

There are isolated intricacies in the computer simulation of a LOCA which do contain some uncertainties. These uncertainties, however, do not preclude reasoned judgments concerning the safety of reactors. Rather, to assure reactor safety it is sufficient to have a conservative accounting of areas of uncertainty so that the overall evaluation or prediction of the accident is conservative.

An evaluation model can thus be thought of as a technical specification of the methods to be used by an analyst in arriving at a conservative assessment of ECCS performance.

In the traditional engineering sense, the evaluation model and the required features of these models specified by the ECCS acceptance criteria provide an acceptable technique for calculating ECCS performance with a factor of safety to accommodate any uncertainty in existing technology.

The required features of evaluation models that are given in the new criteria are more specific than those of the 1971 interim acceptance criteria. The increased specificity is due largely to increased awareness of the subleties of the response of a reactor system to LOCA. This awareness grows with research in LOCA technology.

MORE STRINGENT RESTRICTIONS

On balance, it is expected that use of evaluation models incorporating the features now required will result in slightly more stringent restrictions on the operation of nuclear powerplants than those restrictions now applied on the basis of the interim acceptance criteria. That is, increased knowledge in the past few years has yielded slightly more pessimistic information than optimistic information about ECCS performance.

The increased scientific solidarity associated with the new evaluation models, as compared to the interim models, indicates the strength of the new information. However, it is important to recognize that the older, less sophisticated models were not too different, at least in terms of their impact on reactor operations and on reactor safety. With this background, I would like to briefly discuss the more important of those required features of the new evaluation models which differ from the previously approved models.

CLADDING SWELLING AND RUPTURE

The Zircaloy cladding of some of the reactor fuel elements in both boiling and pressurized reactors will swell and burst if the tempera

ture and pressure reach certain threshold conditions during a LOCA. The new evaluation models are required to explicitly predict swelling and rupture and to account for their effects.

We expect the accounting to have significant effects on several of the important energy transfer mechanisms occurring during a LOCA. Most of these will act to increase the calculated cladding temperature.

Since one of the limiting criteria specifies a maximum allowable cladding temperature, the net effect of a more detailed accounting for swelling and rupture is expected to be a decrease in the maximum allowable heat generation rate for the fuel in some reactors. This may or may not result in the derating of power for a particular plant, depending upon the operating characteristics of the reactor.

PWR CORE FLOW DURING BLOWDOWN

The cores of all modern pressurized water reactors are of the socalled open lattice design. That is, there are no flow restrictions between assemblies of fuel elements, and water is free to move laterally in the core. During a LOCA, water would tend to flow away from higher power assemblies toward cooler portions of the core. This phenomenon was previously accounted for by an arbitrary reduction factor of 0.8 applied to the calculated core flow.

The new evaluation models are required to contain an explicit accounting of flow blockage in hot regions of the core, and thus to provide a realistic estimate of the transient core flow redistribution. The net effect of this change is expected to be an increase in the calculated hot region flow during blowdown with a consequent reduction in the calculated cladding temperature.

PWR REFLOODING CALCULATION

The basic heat transfer data for the reflooding period are those obtained in the FLECHT experiments, sponsored by the Commission as part of its R. & D. program. The validity of these data was reaffirmed, but the new evaluation models do not contain the so-called FLECHT correlations for heat transfer coefficients derived from some of these data to describe cooling of a pressurized water reactor core following reflood by the ECCS.

These previously acceptable correlations have been shown to be inapplicable in several areas, especially for those reactor plants that have low calculated postaccident containment pressures and low reflooding rates.

The new models are required to use the heat transfer data, as opposed to the correlations from the FLECHT tests, that are directly applicable to each particular plant. In addition, the new models are required to assume a degraded heat transfer regime of steam convection whenever the core reflooding rate is calculated to be below a value of 1 inch per second.