Experimental results obtained at integral test facilities (ITF) are used to validate system codes for later application to similar conditions occurring in actual nuclear power plants (NPP). However, scaling and detailed geometry effects play an important role in this process. Even though the code is able to simulate correctly one phenomenon in a determined scale and for a given geometry, it is difficult to state that the code response will maintain the same level of accuracy for a slightly different system. Therefore, it is crucial to experimentally represent the same phenomenology in more than one ITF to assure the capabilities of the employed system code in different geometries and scales.
During 2011, a link has been established between two OECD/NEA projects featuring two different ITF experiments: the PKL test facility and the ROSA Large Scale Test Facility (LSTF). The former is a test facility simulating a pressurized water reactor (PWR) 4-loop Siemens design operated by AREVA and the latter is operated by the Japan Atomic Energy Agency (JAEA) and represents a 4-loop PWR with a Westinghouse design. A counterpart experiments between the two facilities has been carried out with the objective of representing the same accident conditions with two different sets of scaling criteria and different local geometries. The counterpart testssimulated SBLOCA in the hot leg concurrent with a loss-of-offsite-power.
Post-test calculations of the two experiments have been performed by making use of the thermal-hydraulic system code TRACE. The dual analysis aimed at assessing the capability of the simulation models in evaluating the core exit temperature (CET) measurement as an adequate signal to initiate a secondary-side depressurization procedure to prevent unacceptable peak cladding temperature (PCT) excursions. The post-test simulations results reported a good agreement with experimental data in terms of system behavior. However, it has been observed that the correct prediction of the CET time evolution (amplitude and shape) was very sensitive to the details of the modeling of the passive heat structures and flow cross-section in the core and upper plenum (UP) regions, with particular attention to the upper core plate (UCP).