From Installed Capacity to Usable Service: A Probabilistic Assessment of Public EV Charging Accessibility in Chengdu

Public electric vehicle charging accessibility depends on whether installed infrastructure can be converted into services that users can locate, reach, afford, and use reliably. This study applies an expert-informed probabilistic assessment framework to Jinjiang District, Chengdu. Nineteen indicators are organised within a policy–economy–society–technology hierarchy, and judgments from 12 professionals are represented as probability allocations across seven ordered categories. Weibull smoothing, fairness–concentration weighting, and quantile-average (QA) aggregation are applied, with discrete pooling, bounded Beta fitting, sensitivity analysis, and resampling used for robustness assessment. The overall mean remains close to 5.2 across optimised QA, equal-expert QA, and direct discrete pooling, while technology and mobility integration is the most consistently high-rated dimension. Robustness analyses indicate that the overall mean is relatively stable across the principal baseline specifications and local parameter perturbations, while selected indicator-level results and the distribution of expert influence are more sensitive. The findings reinforce the distinction between installed charging capacity and usable service. Public charging planning should therefore move beyond charger counts and distinguish capacity shortages from operational and service-use bottlenecks, with greater attention to charger uptime, information accuracy, site access, queues, payment, pricing, and local coordination.