Comparison of Different Damage-to-Loss Procedures Applied to Site-specific Reinforced Concrete Moment-Resisting Frame Structures

The state-of-the-art probabilistic seismic risk assessment for a given building located at a specific site is summarized in the FEMA P-58 document (FEMA, 2018). This site-specific and building-specific framework follows the PEER-PBEE loss estimation methodology (SEAOC, 1995) which involves five main steps: (i) the probabilistic seismic hazard analysis (PSHA) of the selected location to obtain the rate of occurrence (or exceedance) of one, or more, intensity measures (IMs) of interest at the site; (ii) the response analysis of the specific structure to obtain a link between the levels of the IMs and the performance gauged by one or more engineering demand parameters (EDPs); (iii) the estimation of building damage states, gauged in terms of EDP threshold values, by using fragility functions derived analytically, empirically or based on engineering judgement; and (iv) the linking of different building damage states to losses by using the so-called consequence functions. When merged, steps (iii) and (iv) jointly produce the so-called vulnerability function for the building under consideration. Finally, the fifth and final step (v) involves the convolution of the site seismic hazard curve with the vulnerability function to produce the so-called loss exceedance curve. The first four steps are subject to epistemic and/or aleatory uncertainty. The main goal of this study is to investigate and critically discuss the epistemic uncertainty in building-specific vulnerability curves and, consequently, on the final loss estimates stemming from the application of different fragility-to-vulnerability procedures available in the literature. To do so, two case-study buildings are used: a 4-storey reinforced concrete moment-resisting frame structure with modern ductile design (high code) and a 5-storey structure representative of old design with low-ductility or brittle behavior (low code). The thesis starts with an introduction of the selected location for the case study buildings, the derivation of earthquake hazard at the site, the description of the structural models that are used and a review of the fragility-to-vulnerability methods that are considered for this study. This introductory part is followed by the explanation of the methodology that adopted for the estimation of earthquake-induced losses. The vulnerability and loss exceedance curves obtained by using different methods are then presented and discussed. Keywords: seismic hazard, average spectral acceleration, component-based loss assessment, vulnerability functions, consequence functions, global response-based loss assessment