Abstract
Process industries, including chemical and petrochemical facilities, rely on interconnected systems, making them highly vulnerable to seismic hazards. Earthquakes can cause catastrophic failures, leading to hazardous material releases, structural damage, and severe economic and environmental consequences. Among critical components, steel liquid storage tanks present significant risks due to the large volumes of toxic and flammable substances they contain. Past seismic events have exposed the limitations of outdated design standards, highlighting the need for robust seismic vulnerability assessments. Fragility curves play a crucial role in seismic risk evaluation, offering probabilistic estimates of damage under varying earthquake intensities. These curves are developed through empirical, expert-based, and analytical methods, with analytical approaches preferred for their systematic treatment of uncertainties. This review critically examines analytical methodologies for deriving fragility curves specific to steel liquid storage tanks. It first analyzes historical earthquake-induced failures, identifying key damage states and failure mechanisms. A comprehensive review of numerical modeling techniques is presented, covering simplified spring-mass models and advanced finite element methods. The paper then explores analytical fragility methodologies, emphasizing uncertainty management and optimal intensity measures for probabilistic seismic demand analysis. Additionally, emerging trends in surrogate model-based fragility analysis are discussed, highlighting machine learning and advanced regression techniques for enhanced computational efficiency. The review concludes by summarizing key findings, identifying research gaps, and offering recommendations to improve seismic vulnerability assessment. These insights aim to enhance risk mitigation strategies, ensuring the resilience and safety of process industry facilities in earthquake-prone regions.
