Optimal management of V2X in parking lots

Jan Engelhardt, Anna Malkova, Xihai Cao, Pietro Zunino, Jan Martin Zepter, Haris Ziras, Mattia Marinelli, Benedikt Baldursson, Maria Hach, Herbert Amezquita, Hugo Morais, Cindy P. Guzman

Research output: Book/ReportReportCommissioned

27 Downloads (Pure)

Abstract

The present deliverable D2.3 presents the modelling framework and research backbone to the development of Vehicle-to-Everything (V2X) strategies for the demonstration activities in Denmark as part of the EV4EU project. The charging infrastructure at both locations of the Danish demo (Risø and Campus Bornholm) is identical and comprises six Alternating Current (AC) chargers with two connectors (dual outlet) providing 11 kW charging power each. The twelve charge points are connected to the grid with a maximum cluster power limit of 43 kW. The developed V2X strategies aim to provide a seamless integration of the charging infrastructure in the electricity system and serve two main purposes. Firstly, they ensure the compliance with the grid connection limit, as the cluster limit of 43 kW is lower than the sum of the rated power of all charge points. Secondly, they utilize the flexibility of charging processes to provide services both behind and in front of the meter.

The modelling framework comprised forecast models to predict the Electric Vehicle (EV) charging demand, as well as the expected generation from local Renewable Energy Sources (RES). The forecast models were integrated in the control models developed within the scope of this work. Moreover, users of the charging infrastructure at the Danish demo sites can provide their estimated departure time and energy needs via an app, which is taken into account in the decision-making process.

The control models facilitate the coordination of EV charging, which offers the possibility to provide services to local and external grids. This is tested through a set of different scenarios and control objectives. The conducted simulations utilize actual Photovoltaic (PV) production and EV charging data from winter and summer weeks in Denmark. It is shown that the proposed control methods achieve the compliance with local grid connection limits and are further able to provide demand response to Distribution System Operators (DSOs) by reducing the cluster power limit temporarily and on request. The tested reduction of the cluster power limit by half from 43 kW to 21.5 kW on weekdays between 7 a.m. and 12 p.m. leads to a decrease of the fulfilled energy requested by the EVs by 5% in the summer scenario and 6% in the winter scenario. Yet, this decrease is small considering that the power reduction is requested for 5 consecutive hours with frequent EV arrivals. The consideration of local PV generation, as present at the demo site Campus Bornholm, allows to cover parts of the EV charging demand through local generation from RES. In the summer scenario, an overall proportion of 77% of the energy delivered to EVs is provided by the local PV system, defining the system’s self-sufficiency. Conversely, 27% of the total energy provided by the local PV system is utilized for EV charging, defining the system’s self-consumption. In the winter scenario, the values for self-sufficiency and self-consumption are 13% and 84%, respectively.

In case single-phase EVs are present in charging clusters, the charging processes can lead to phase unbalances. To counteract this effect, the present work proposes an approach for including phase unbalance constraints in the optimization problem formulation and demonstrates its functionality through simulations. Moreover, the technical capability of EV charging infrastructure to provide frequency control was assessed. The work investigated a distributed control method utilizing the autonomous controllability of the chargers deployed in the Danish demonstration. The model parameters, such as delays from hardware and software, were characterized through the installation at Risø Campus. The simulation results showed that the distributed control architecture enables on average a 9% faster response compared to the centralized method and followed the reference value more closely.

The developed control methods serve as a basis for the upcoming demonstrations in Work Package 9 (WP9). In this regard, the experimental testing is expected to provide important insights on how the proposed methods perform within real operational settings.
Original languageEnglish
PublisherEuropean Climate, Infrastructure and Environment Executive Agency
Number of pages47
Publication statusPublished - 2024

Fingerprint

Dive into the research topics of 'Optimal management of V2X in parking lots'. Together they form a unique fingerprint.

Cite this