Effects of Multiple Uncertainties on the Seismic Fragility of High-Voltage Porcelain Column-Type Equipment Systems with Geometrically Symmetric Components

High-voltage porcelain column-type equipment systems comprise various equipment units whose porcelain columns generally have circular cross-sections and approximately axisymmetric geometries. Although this geometric symmetry results in nominally equivalent lateral mechanical properties at the component level, differences among equipment types and multiple uncertainty sources may lead to heterogeneous seismic responses at the system level. To investigate the combined effects of ground-motion randomness and uncertainties in the elastic modulus and diameter of porcelain sleeves and damage control indices, Latin hypercube sampling was employed to generate 100 structural models with different modeling parameters. These models were randomly paired one-to-one with 100 selected ground-motion records, and uncertainty in the damage control indices was incorporated into the dynamic response and fragility analyses. The results indicate that the variability in structural seismic responses arises from the combined effects of modeling-parameter uncertainty and ground-motion randomness. These effects are propagated to the fragility curves through changes in the median ground-motion intensity and total logarithmic standard deviation of the fragility functions. The fragility curves accounting for multiple uncertainty sources generally fluctuate around the baseline curves considering ground-motion randomness alone, without exhibiting a consistent upward or downward shift. No strictly monotonic relationship was observed between the coefficients of variation in the uncertain parameters and the total dispersion of the fragility results, although larger parameter variability produced greater deviations from the baseline in some cases. Nevertheless, the overall differences remained limited, with a maximum absolute difference of 0.043. From the perspective of symmetry, the results demonstrate that component-level geometric symmetry coexists with system-level response heterogeneity and cannot alone eliminate the effects of equipment-specific characteristics and stochastic uncertainties on seismic fragility. Within the scope of this study, fragility curves considering ground-motion randomness alone may therefore provide a reasonable approximation for engineering assessment.