Modelling of electric vehicle user profiles for flexibility management and charging infrastructure planning

Full Text
embargat.txt embargoed access
Request a copy
When filling the form you are requesting a copy of the article, that is deposited in the institutional repository (DUGiDocs), at the autor or main autor of the article. It will be the same author who decides to give a copy of the document to the person who requests it, if it considers it appropriate. In any case, the UdG Library doesn’t take part in this process because it is not authorized to provide restricted articles.
Share
The transition to electric mobility is facing multiple challenges, usually associated with the roll-out of the charging infrastructure. On one hand, cities must develop a charging infrastructure that meets the user needs while the type of electric vehicle (EV) users is specific for every charging area. On the other hand, a high EV demand in the power system can bring congestion issues at the low-voltage power grid and this can involve power supply quality issues and a barrier to further development of the charging infrastructure. This thesis aims to provide tools to solve the challenges raised in both stages of the adoption of electric mobility. With this purpose, a methodology to cluster and model generic EV user profiles based on connection patterns is proposed and applied to these two key areas: flexibility management and charging infrastructure planning. The concept of user profiles is introduced as a tool to identify common connection patterns with a characteristic flexibility potential. A clustering methodology using Gaussian Mixture Models (GMM) is applied based on variables such as connection start time and duration. Common usage patterns in public charging infrastructure are observed, providing insights into EV user behaviour. The profiling methodology is validated with three different real data sets of charging sessions along with the three journal articles that shape the core of this thesis. The clustering methodology is followed by a modelling methodology to perform stochastic simulations of EV charging sessions in terms of connection times, required energy and charging power rate. Modelling every user profile independently lets to simulate a wide range of scenarios since the share of each user profile over the total EV demand can be configured according to the environment (i.e. location, time horizon, etc.). This application is explored with two journal articles where scenarios with high penetration of EV sessions are simulated to (1) optimally size a charging hub and (2) analyse the required number of charging points of city-level charging infrastructure. In both contributions, the configuration of user profiles in specific areas is crucial for properly sizing charging infrastructure, avoiding extra costs that harm the business model or losing EV users’ confidence with undersized installations. This thesis also compares different smart charging strategies through simulations, as well as the benefits that the user-profile approach could bring to smart charging programs. When scheduling individual sessions according to an aggregated demand setpoint, the extra knowledge of profiling EV users beforehand can provide insights for a more reliable flexibility prediction. Moreover, scheduling sessions from selected user profiles could lead to exploitation cost savings and reduced impact on EV users. Finally, the application of a smart charging program at the city level with high penetration of EVs has been also simulated to analyse its impact on all stakeholders involved in the EV charging sector, from the final EV user to the charging operator business model. Curtailing charging power based on dynamic capacity signals proves effective to avoid grid congestion and defer reinforcements of the existing power grid while expanding the charging infrastructure and supplying the majority of the energy required by EV users. Overall, this thesis enhances understanding of EV user behaviour, analyses different smart charging strategies, and provides insights for charging infrastructure planning. These findings have practical implications for stakeholders involved in the EV ecosystem, contributing to the ongoing transition to electric mobility. ​
​L'accés als continguts d'aquesta tesi queda condicionat a l'acceptació de les condicions d'ús establertes per la següent llicència Creative Commons: http://creativecommons.org/licenses/by-nc-sa/4.0/