ENSI-AN-9657

ENSI closely followed the PRP project from its inception with an experienced expert team. In its final review report ENSI acknowledged that the state-of-the-art in probabilistic seismic hazard assessment was further improved by the PRP project. ENSI assessed the achieved refinements in the project focal points – the «ground motion characterization» (subproject 2) and the «site response characterization» (subproject 3) – to be well-founded. In contrast, the «seismic source characterization » (subproject 1) was not investigated in sufficient detail, according to ENSI. After it became evident late in the project that the model modifications in subproject 1 had a significant influence on the computed seismic hazard, the experts did not have the opportunity to question or to confirm their assessments anymore. The «seismic hazard computation» (subproject 4) was conducted in an appropriate manner and the applied software met the accepted specification. Nevertheless, due to the concerns regarding subproject 1, ENSI could not accept the final results of the PRP.

Due to the reservations concerning PRP subproject 1, ENSI initiated a sensitivity analysis in which the model for subproject 1 was replaced by the corresponding model of the Swiss Seismological Service (SED). The results of this combined «SED-PRP model» were found to be higher than the results of both the PRP and the SED model. In May 2016, ENSI ordered the implementation of the results of the «SED-PRP model», denoted as seismic hazard assumptions ENSI-2015 (in German «Erdbebengefährdungsannahmen ENSI-2015»). At the same time ENSI required the NPP operators to assess the consequences on the safety of the nuclear power plant and, in particular, on the risk. To this end the NPP operators must update the following in respect of earthquakes: a) by the end of 2018 the safety cases originally required by ENSI after the Fukushima reactor catastrophe in March 2011, b) by mid-2019 the probabilistic safety analysis, and c) by the end of September 2020 the deterministic safety analysis.