A Digital Sandbox Approach: Simulating and Forecasting Charging Demand of Electric Two-Wheelers for Risk-Informed Infrastructure Planning

The rapid surge of Electric Two-Wheelers (E2Ws) in high-density urban villages imposes severe strain on low-voltage residential distribution networks. Unlike formal Electric Vehicles, E2W charging is decentralized and highly constrained by short pedestrian walking thresholds, frequently forcing users to adopt non-compliant “fly-wire charging” when public facilities are scarce. Traditional top-down load models fail to capture these localized, micro-behavioral single-phase grid impacts. To address this deficit, this study proposes a GIS-integrated “Digital Sandbox” simulation framework that projects individual behavioral mutations directly onto feeder networks via a hyper-granular “particle tracking” mechanism, treating each E2W as an autonomous agent. As a case study focused on a representative urban-village area, we validate the model using field data from the site. A 30-day simulation reveals that unmanaged fly-wire charging generates a peak load of 17.64 kW (nearly double the public station peak) and accounts for 38.9% of aggregate energy consumption—concentrated within the top 10 buildings and coinciding with evening peaks, inducing severe phase imbalance. While the numerical results are case-specific, the framework itself is transferable to other service areas through re-calibration against local data. This foundational digital twin blueprint shifts E2W planning from guesswork to particle-level risk prediction.