Abstract

Electric-vehicle charging can erode operating margins in distribution networks, yet impacts are often localized rather than system-wide. This study quantifies unmanaged charging effects on the Oujda 60/22-kV system using a geo-referenced steady-state power-flow model coupled with a stochastic charging generator for cars, motorcycles and buses. The network representation comprises 124 substations and 117 branches, and we examine snapshots at 0%, 10% and 30% adoption. In thermal terms, the bulk of the network remains comfortably loaded: the share of lines operating at ≤70% of their rating is 81% at baseline, 79% at 10% and 78% at 30%. Localized constraints intensify modestly with penetration: the share of overloaded lines (>100%) rises from 6% to 7% and 9%, and the worst-loaded span increases from 151.6% to 157.2% and 168.1%. Voltage performance is similarly robust in bulk (median around 0.97 per unit), with a small weak-bus tail near charging hotspots. All cases converged reliably. The workflow is lightweight and reproducible, supporting feeder-level hosting-capacity screening and motivating targeted reinforcement or simple smart-charging measures in data-constrained systems.