Enlargement of the uncertainty database of the CIAU methodology

Best-estimate thermal-hydraulic system code have been developed and are currently used to simulate transients behaviour in small scale test facilities and in NPPs. While the error made in predicting plant behaviour is called uncertainty and it is unknown, the discrepancy between measured and calculated trends related to tests performed in experimental facilities is defined as the accuracy of the predictions and constitutes the known error. As a consequence, the process of application of best-estimate (or realistic) computer codes to the safety analysis of NPPs implies the evaluation of uncertainties. 'Sources of uncertainty' affect the prediction of the results of best-estimate codes and must be taken into account. In this context, the Code with (the capability of) Internal Assessment of Uncertainty (CIAU) has been proposed by the University of Pisa to realize the integration between a qualified system code and an uncertainty methodology and to supply proper uncertainty bands each time a NPP transient scenario is calculated. At the basis of the derivation of CIAU, there is the UMAE (Uncertainty Methodology based on Accuracy Extrapolation) previously proposed by University of Pisa and the RELAP5/MOD3.2 system code. The CIAU method has been recently extended to evaluate the uncertainty of coupled 3-D neutron-kinetics/thermal-hydraulics calculations. The result is CIAU-TN. The feasibility of the approach has been demonstrated and sample results related to the turbine trip transient in the Peach Bottom NPP and to the OECD/NRC PWR MSLB Benchmark have been obtained. The paper mainly discusses about the enlargement of the CIAU (uncertainty) database obtained including 11 new tests concerning different transients and facilities. The process by which new qualified tests (in the sense required by UMAE methodology) are added to the existing database constitutes the current main priority of the CIAU methodology and it is continuously ongoing. The 'internal' qualification process of the new database has been fulfilled and the application of the updated uncertainty matrices on the CL-18 OECD/CSNI UMS test (SBLOCA in the LSFT facility) has been also considered in order to compare the new uncertainty bands with the ones previously calculated and to demonstrate the 'external' (independent) qualification of the CIAU database. The use of best-estimate codes within the reactor technology, either for design or safety purposes, implies understanding and accepting the limitations and the deficiencies of those codes. Therefore, uncertainty statements must supplement the application of best estimate codes. A method to calculate the uncertainty associated with NPP computer code calculations directly integrated in the code has been briefly presented. CIAU constitutes a powerful tool that is originated by the combination of a qualified best-estimate system code and a suitable uncertainty methodology. The implementation of the CIAU capability allows the achievement of error (uncertainty) bands coupled with the results of the concerned system code calculation. The main advantage of an IAU approach consists in avoiding, from the methodology user point of view, to interpret logical statements that are part of the application process for all current uncertainty methods, i.e. avoiding user effect when using uncertainty methodologies. The above consideration does not exclude the use of engineering judgment: rather, engineering judgment is embedded into the development of the IAU method and it is not needed in its application