Use of separate-effects experiments in verification of system thermal-hydraulics

In recent years, a number of advanced, best-estimate systems codes such as TRAC and RELAP5 have been developed in order to accurately predict the consequences of various postulated accidents and transients in Light Water Reactor (LWR) systems. Although these codes had to go through some verification or assessment during the developmental stage, it has been recognized that an independent assessment of these codes is necessary before they should be applied to any decision making process. The USNRC is, therefore, sponsoring such efforts at several national laboratories including BNL. The overall assessment matrix includes separate-effects, integral and plant tests. However, this paper will focus on how the separate-effects tests can be utilized in verifying the thermal-hydraulic models that control the various stages of postulated accidents and/or transients in a LWR system.