| Abstract: | A technique was developed to evaluate the applicability of data from small scale facilities for validation of codes for analysis of nuclear safety with emphasis on the next generation of reactors. The technique first divides an accident into phases based on the components that come into play as the accident evolves. Conservation equations, resolved to the component level and their interconnections, are derived for the active components in each phase. The equations are then nondimensionalized and reference parameters are selected such that the dependent variables, other than the system response of interest, are of order 1. Order of magnitude analysis is then performed for each equation and then between equations, based on the numerical values of the nondimensional coefficients for each term, with only the large order terms being retained. The resulting equations then contain terms whose impact on key system responses (e.g. reactor vessel level) are ordered in terms of the magnitude of the nondimensional groups multiplying the O1] dependent variables. The reduced set of equations and nondimensional groups are validated with experimental data where possible. The validation process is meant to demonstrate that the important terms have been retained and enhance confidence in the system of equations used to capture the main processes occurring in each phase. The methodology was demonstrated by evaluating the applicability of small-scale facility data for next generation reactor SBLOCA. Based on the nondimensional equations, the dominant nondimensional groups, and hence the dominant physical mechanisms and their dependence on geometric and operational parameters, were identified for a particular scenario, an AP600 cold leg break, starting from the initiating event through long term cooling. The important parameters entering the groups included elevation differences between the reactor vessel and other components, PRHR heat transfer rates, fluid thermophysical properties, liquid levels in tanks, flow resistances in the CMT lines and IRWST lines, flow resistance in the pressurizer surge line, and pressurizer drain rate. It was also shown that, after the beginning of CMT draining and accumulator injection, the dominant processes do not depend on break size provided they are small. The dominant processes were dependent on plant geometry and the operation of engineered safety features, such as the automatic depressurization system. The same transient events were evaluated for three experimental facilities and the same nondimensional groups, and hence mechanisms, were shown to be important. It was found that these nondimensional groups covered the range expected in the AP600, indicating that while there may be some distortions in scaling for a particular facility, between them, the important phenomena were captured and the small-scale facility data appear applicable for SBLOCA in the AP600 system. In more general terms, the methodology appears suitable for assessing scaling of various facilities for other postulated accidents and for other reactor concepts. |
