Seismic acceleration demand and fragility assessment of storage tanks installed in industrial steel moment-resisting frame structures

Recent seismic events highlighted the importance of non-structural elements in seismic assessment and performance-based earthquake engineering. When referring to industrial facilities, storage tanks represent one of the most critical components for the continuous functionality of industrial and building plants after seismic events. At the same time, the dispersion of dangerous or inflammable materials, due to their collapse, could become a safety issue in terms of both human life and environmental impact. This study focuses on the seismic performance assessment of a liquid-storage tank installed in an industrial steel moment-resisting frame building. The performance is assessed in terms of fragility curves, that are provided for the meaningful limit states that affect the liquid-storage tank and considering several intensity measures. The results related to the structural performance are also discussed in detail in terms of fragility functions and floor response. The assessment is performed through nonlinear dynamic analyses using two different modelling approaches to evaluate the structural and non-structural performance. The first modelling approach explicitly accounts for the tank-structure interaction, while in the second one the tank is only modelled considering its seismic mass. The results show the importance of the modelling assumptions to accurately estimate the seismic demand on liquid-storage tanks and corresponding fragility, which is calculated in terms of different intensity measures (IM): average ground spectral acceleration, peak floor acceleration, and floor spectral accelerations. Peak floor acceleration, drift profiles, and absolute acceleration and relative displacement floor response spectra are instead scrutinised for what concerns demand. The comparison with the code formulations available in the literature for estimating the seismic demand to non-structural elements pointed out their inaccuracy for the analysed liquid-storage tank and the need for further efforts to develop more accurate procedures. Furthermore, the outcomes indicate that the IM adopted for secondary components, i.e. peak floor acceleration, may not always be the optimal one hence additional attention should also be paid to the definition of updated fragility curves for non-structural elements, identifying and adopting more suitable IMs.