Customer load modeling is a fundamental component of many business practices in the energy industry, particularly in the electricity distribution and retail sectors. The increasing numbers of electric vehicles (EVs) and thereby the increase in EV charging loads connected to the grids have significant implications for distribution system operators (DSOs) as they may result in increased peak loads, major changes in load profiles, and a need to invest in the infrastructure. It is also worth considering the potential opportunities that dynamic EV charging loads can bring, for instance, for the demand-side management.

The main focus of this doctoral dissertation is to investigate and provide a broad but also detailed overview of the load modeling of EV charging (spatial and temporal) mainly from the distribution system operator’s (DSO’s) business perspective, but also to cover the perspectives of the other players, such as the electricity retailer and the aggregator.

The main objective of the doctoral dissertation is to offer tools to support and improve the DSO’s long-term strategic planning process in order to identify challenges or opportunities of the large-scale EV smart charging. Modeling tools and analysis of EV charging loads also provide important information for other players in the field of energy systems and energy markets. The main focus is on determining the stochastic nature of the charging loads to facilitate load formation and forecasting activities. Spatiotemporal modeling techniques are also applicable for analyzing and managing electricity retailers’ or flexibility aggregators’ customer portfolios and (load) profile risk.

As the main outcome, the doctoral dissertation shows the variety of EV charging applications and modeling practices. The novelty of the dissertation is in summarizing a wide range of EV charging applications and laboratory experiments to quantify the impacts of EVs on the power system and its various parties. A particular novelty lies in adding the features of the cold environment to the load modeling of EVs, but also in showing that spatial modeling benefits from using convolutional neural network (CNN) models. As the main conclusion, it can be stated that EV charging will have a significant impact on the power system, but the impact will depend on the development of smart charging applications.