Seismic fragility analysis of a coupled tank-piping system based on artificial ground motions and surrogate modeling

Abstract

The catastrophic consequences of recent NaTech events triggered by earthquakes highlighted the inadequacy of standard approaches to seismic risk assessment of chemical process plants. To date, the risk assessment of such facilities mainly relies on historical data and focuses on uncoupled process components. As a consequence, the dynamic interaction between process equipment is neglected. In response to this gap, researchers started a progressive integration of the Pacific Earthquake Engineering Research Center (PEER) Performance-Based Earthquake Engineering (PBEE) risk assessment framework. However, a few limitations still prevent a systematic implementation of this framework to chemical process plants. The most significant are: i) the computational cost of system-level simulations accounting for coupling between process equipment; ii) the experimental cost for component-level model validation; iii) a reduced number of hazard-consistent site-specific ground motion records for time history analyses.In response to these challenges, this paper proposes a recently developed uncertainty quantification-based framework to perform seismic fragility assessments of chemical process plants. The framework employs three key elements: i) a stochastic ground-motion model to supplement scarcity of real records; ii) surrogate modeling to reduce the computational cost of system-level simulations; iii) a component-level model validation based on cost-effective hybrid simulation tests. In order to demonstrate the potential of the framework, two fragility functions are computed for a pipe elbow of a coupled tank-piping system.

Gaussian process regression surrogate model for dynamic analysis to account for uncertainties in seismic loading

“…Several previous studies have addressed uncertainties in dynamic seismic loading. For example, studies by Sainct et al 6 and Abbiati et al 7 addressed seismic load uncertainties by using seismic model parameters and seismic strength indices as inputs. However, the problem is a discrepancy between the observed seismic waveform and the waveform generated from the seismic model.…”

Multi‐hazard fragility assessment of monopile offshore wind turbines under earthquake, wind and wave loads

“…Inspired by the dual-intensity-measure (IM) vulnerability assessment adopted for the case of coupled tank-piping systems, 22 buildings [23][24][25][26] and bridges, [27][28][29] there has been a recent extrapolation of knowledge to the field of multi-hazard reliability assessment for OWTs, with the aim to produce fragility surfaces. To realistically assess risks in a probabilistic manner, it is necessary to accurately capture the simultaneous dynamic loads and the response of the OWT structure when subjected to these loads.…”

Seismic Fragility Curves of Steel Frames Equipped with Easily Repairable Dissipative Devices

“…In recent years, the probabilistic Performance-Based Earthquake Engineering (PBEE) methodology has become popular, and an increasing number of seismic fragility functions for a particular structural typology or component, crucial for both risk assessment and/or a probabilistic PBEE application, are being developed for both civil [1] and industrial buildings [2] [3]. On these premises, in this paper Incremental Dynamic Analyses (IDAs) were applied to investigate the influence of equipping steel frames with two different dissipative seismic devices analysed in DISSIPABLE, namely the Dissipative Replaceable Beam Splice (DRBeS) and the Dissipative Replaceable Link Frame (DRLF).…”