| Title |
Field Measurement Based Estimation and Expansion of Hourly Charging Patterns of Personal Slow EV Chargers on Jeju Island |
| Authors |
김수완(SuWan KIM) ; 이개명(Gae-Myoung Lee) ; 김호찬(Ho-Chan Kim) |
| DOI |
https://doi.org/10.5370/KIEE.2026.75.10.2813 |
| Keywords |
Electric vehicle charging; Hourly load decomposition; Personal slow charger; Measurement validation; Time-of-use tariff |
| Abstract |
Personal slow chargers account for 83.4 % of all EV chargers on Jeju Island, yet their hourly charging behavior is unobservable because these units have no communication capability and utility metering records only monthly energy by time of use (TOU) period. This paper presents a field measurement based framework that estimates their hourly charging profile and expands it toward the population. From nine months of charging protocol logs and metering records of five 7 kW class residential chargers, 832 sessions and 5-minute power trajectories were reconstructed. The reconstruction was verified along three independent paths: monthly aggregates matched utility meter readings; simultaneous inlet outlet measurement showed a 0.54 % mean deviation; and the resulting TOU composition agreed with metered records within 0.5 percentage points. Leave one out cross-validation, in which each unit was estimated from the most similar remaining unit, yielded an energy weighted mean absolute error of 2.70 percentage points. Using hourly allocation ratios derived by season and day type, 3 036 chargers with similar TOU composition were selected from 15 984 quality filtered candidates and their monthly energies decomposed into hourly profiles while preserving monthly totals. Both the measured units and the expanded group show late night concentration, with 62.2 % and 77.9 % of energy consumed between 22:00 and 06:00, while only 6.8 % falls within the daytime hours when photovoltaic surplus arises. The results provide an hourly demand basis for identifying chargers that could absorb surplus generation without hardware replacement. |