Python Supervised Co-simulation for A Day-long Harmonic Evaluation of EV Charging

Abstract

To accurately simulate electric vehicle DC fast chargers' (DCFCs') harmonic emission, a small time step, i.e., typically smaller than 10 μs, is required owing to switching dynamics. However, in practice, harmonics should be continuously assessed with a long duration, e.g., a day. A trade-off between accuracy and time efficiency thus exists. To address this issue, a multi-time scale modeling framework of fast-charging stations (FCSs) is proposed. In the presented framework, the DCFCs' input impedance and harmonic current emission in the ideal grid condition, that is, zero grid impedance and no background harmonic voltage, are obtained based on a converter switching model with a small timescale simulation. Since a DCFC's input impedance and harmonic current source are functions of the DCFC's load, the input impedance and harmonic emission at different loads are obtained. Thereafter, they are used in the fast-charging charging station modeling, where the DCFCs are simplified as Norton equivalent circuits. In the station level simulation, a large time step, i.e., one minute, is used because the DCFCs' operating power can be assumed as a constant over a minute. With this co-simulation, the FCSs' long-term power quality performance can be assessed time-efficiently, without losing much accuracy.