Stochastic Load Modeling of High-Power Electric Vehicle Charging - A Norwegian Case Study

Abstract

In recent years, the number of electric vehicles (EVs) has increased rapidly. Due to technological advancement, government policies and the focus on reducing greenhouse gas emission, the growth can be expected to continue. Home charging of EVs will often be sufficient for short-distance travel and daily routines. However, EVs still have a limited range. Thus, for longdistance travel, a network of fast charging stations (FCS) is needed. The stochastic nature, high power demand and short duration of EV fast charging, make it in many cases a grid capacity issue rather than an energy issue. Therefore, knowledge about the load profile of FCSs is important. In this paper, a model is developed for the simulation of the aggregated load profile of an FCS. The FCS load model includes a mobility model based on actual traffic flow, EV charging curves and temperaturedependent EV efficiency. Simulations are performed using the Monte Carlo simulation technique, to get a daily load profile for the FCS. Real-world data for the studied FCS in Norway is compared with the results from the simulation to analyze the performance of the FCS load model. The developed load profile for the FCS has a high peak-to-average power ratio, which indicates that the socioeconomic profitability of fast charging stations still is low.

Heavy-duty electric vehicle charging profile generation method for grid impact analysis

“…To generate load profiles for an HPCS for HDEVs, a method for modelling the aggregated load of the HPCS is developed. The proposed method is based on the work performed in [7] which uses a stochastic approach for passenger EVs. Due to less known specifications of HDEVs and their behaviour compared to state-of-the-art electric cars, some simplifications have been made.…”

“…The time between each HDEV arriving the HPCS, in minutes, is calculated with (2) as described in [7], where u is a variable that is uniformly distributed between 0 and 1.…”

“…The HDEV will either charge until the charging session has lasted as long as the mandatory break, T break , (7) or when the maximum charged energy is reached (8).…”

“…A second and more common method is to build the model on stochastic parameters without the autonomous decisionmaking and interactions between the agents. References [6] and [7] are using Poisson processes with a predefined arrival rate or historical traffic flow data directly. The initial state of charge (SOC) is drawn from various probability distributions in [6]- [8].…”

“…Instead, data from HPCSs in Norway and Sweden containing charging duration is used. Monte Carlo simulations are often applied to evaluate the uncertainty of the stochastic parameters [3], [6], [7].…”