PSI - Issue 78

Yazdan Almasi et al. / Procedia Structural Integrity 78 (2026) 433–440

437

or momentum flux (Basco & Salzano, 2017; Nishino et al., 2024; Park & Cox, 2016). However, the scarcity of such data limits its reliability for industrial assets. Therefore, analytical fragility models, often based on simplified force balance equations and hydrostatic pressure distributions, are used to simulate uplift, sliding, and overturning (Araki et al., 2017; Basco & Salzano, 2017; Mebarki et al., 2016). For atmospheric storage tanks, anchorage and fill level again emerge as the most influential parameters, with uplift thresholds defined by buoyancy-over-weight ratios and foundation cohesion (Araki et al., 2017; Basco & Salzano, 2017; Mebarki et al., 2016). Recent studies have also employed simulations or coupled structural-hydrodynamic models to estimate better force distributions and the timing of impact (Goda & De Risi, 2023; Nishino et al., 2024). These models enable the identification of limit states such as anchorage failure, tank shell rupture, or foundation scour, and are used to generate fragility curves in terms of flow depth or velocity thresholds. While single-hazard fragility curves remain the dominant practice, several efforts have emerged to capture cascading effects, particularly the degradation of structural resistance following an earthquake and prior to tsunami impact. Sequential analysis frameworks, where a structure’s post-earthquake residual capacity is used as the initial condition for tsunami loading, have been proposed, though widespread implementation remains limited due to computational complexity and data demands (Goda & De Risi, 2023; He et al., 2022; Wu et al., 2024).

Fig. 2. (a) A fuel tank in Kesennuma City destroyed by the 2011 GEJET; (b) Oil spill fires at Kesennuma Bay in the 2011 GEJET; (c) Heavy oil tank at a thermal power plant destroyed by the tsunami in 2011. (Source: (a), (b) Nishino and Imazu (2018); (c) Krausmann et al. (2019)) Accurate risk quantification also requires understanding equipment exposure and potential consequences. Industrial exposure is characterized by facility layout, elevation, and protective systems. GIS-based tools are employed to assess spatial vulnerability, particularly in coastal and high-risk zones. Consequences such as toxic releases, fires, and operational disruptions are influenced by equipment proximity and interconnectivity. Integrated risk indicators, including expected annual loss, inform retrofitting priorities and resilience-based design in Na-Tech-prone areas (He et al., 2022; Krausmann et al., 2017, 2019). Table 1. Fragility Modelling Approaches for Industrial Equipment under Earthquake and Tsunami Hazards

Hazard

Equipment

EDPs

Limit States

Modelling Methods

PGA, PGV, Sa, Strain thresholds, Internal pressure (for pipelines and vessels) Flow depth, Flow velocity, Momentum flux, Impulse load, Buoyancy-weight ratio, Foundation cohesion

Anchorage failure, shell buckling, roof sloshing, uplift, nozzle tearing, pipe joint separation, rupture, sliding Shell buckling, uplift due to buoyancy, sliding, overturning, anchorage failure, shell rupture, foundation scour

NLTH analysis, MSA, Beam-shell FE models, Segment-based or pseudo-static models, Monte Carlo simulations, Lognormal cumulative distribution fits Empirical, Analytical, Hybrid, Coupled structural-hydrodynamic simulations, Energy-Flux Models, Wave Impact Analysis, Weibull-based Models, Nonlinear Static FE

ASTs, Pressure vessels, Pipelines, Horizontal vessels

Earthquake

Tsunami

ASTs, Pipelines

4. Multi-hazard risk assessment methods The growing recognition of the complex interactions between natural hazards has spurred research into multi-hazard risk assessment methods, particularly for critical infrastructures such as industrial facilities. In coastal industrial zones,