Fire Fragility Curves for Industrial Steel Pipe-Racks Integrating Demand and Capacity Uncertainties

Abstract

This paper aims at deriving fire fragility curves for a prototype steel pipe-rack in an industrial plant subjected to localised fires. In particular, starting from a reference case study, uncertainties related to the structural capacity and the size of the localised fires caused by a hole in a tank or a hole in a pipe are included in the analyses. Thus, the influence of uncertainties in the derivation of the fragility functions was highlighted by comparing four sets of analyses in which both demand and capacity uncertainties were progressively introduced. Moreover, alongside the cloud analysis (CA), the suitability of the Multiple Stripe Analysis (MSA) to build relevant probabilistic fire demand models was assessed. Fire fragility curves were derived by considering the interstorey drift ratio (ISDR) as engineering demand parameter (EDP) and by assessing different relevant intensity measures (IMs) that represent the severity of localised fires. It was found that by introducing uncertainties in the steel yield strength, lower probabilities to exceed the life safety and the near collapse limit states with respect to the reference case study were observed. Moreover, the inclusion of further uncertainties, described with continuous physically-based probability functions of the size of the fire diameter, affected the probabilistic models by lowering the probability of exceedance. These functions provide a more realistic description of the fire scenario, enabling a better representation of the structural vulnerability. For this case study, the CA exhibited better suitability for the derivation of fire fragility curves than the MSA. All the analysis results are thoroughly discussed in the paper.