A semi-analytical formulation for accounting uncertainties of hazard parameters in structural seismic reliability analysis

The probabilistic seismic hazard analysis (PSHA), is currently one of the most used approaches worldwide for assessing seismic hazard, and represents the mostly used approach adopted for the development of seismic maps. PSHA relies on strong mathematical bases, and it is a correct application of the Total Probability Theorem; it is thus able to combine three main sources of uncertainty, e.g. the earthquake magnitude, the source-to-site distance and the corresponding ground-shaking scenario. As a consequence, because of its intrinsic nature, also model parameters can be sources of variability, since most of the time they are extrapolated from historical data. Thus, this work wants to give a contribution on the debated problem of uncertainty in seismic hazard estimates, by proposing a semi-analytical formulation able to include uncertainties arising from model parameters, treating them, in turn, as random variables. The proposed formulation adopts the reliability index and its standard deviation for computing hazard curves characterized by an assumed probability to be underestimated. In the second part of the work, the formulation is applied to a case study represented by an existing bridge, showing its practical use and investigating how different levels of knowledge of seismic hazard model input parameters, can impact the outcomes of a classic structural seismic reliability or risk analysis carried out without taking into account such specific issue.