Congestion Management of EV Charging in Distribution Networks

Sammanfattning: The societal integration of electric vehicles (EVs) imposes several challenges on the electrical grid, where congestion and overcurrents through related components are of focus for this thesis. Active network management through control systems is one flexible solution to this problem, and in the project Active Network Management for All (ANM4L), a congestion management algorithm utilizing PI controllers has been developed. It has the possibility to control the active power (P) of the network by curtailing the charging power for the active EVs when overcurrents are detected. The focus of this paper is to evaluate the effectiveness of the ANM algorithms performance in achieving congestion management, by comparing with an uncontrolled scheduling and a decentralized tariff-based scheduling. The purpose of the latter is to investigate the potential for congestion management without aggregator involvement and to minimize the charging costs for the EV owners by scheduling all charging to low-cost hours. The test network was designed without consideration of extensive EV charging, illustrating a present day network which might have been dimensioned decades ago. The maximum number of actively allowed EV units in the network was led by the ANM controlled scheduling at 53%, followed by the uncontrolled (37%) and tariff-based (24%) scheduling. This illustrates the future network and societal constraint in terms of EV integration. For the ANM implementation, two prioritization schemes were implemented. In a 10-bus low voltage test network with 11 kW home-charging stations, the ANM algorithm proved to be efficient in alleviating the network constraints whilst maintaining the EV owners energy demands when owners with the highest instantaneus power consumption experienced the highest curtailment. The tariff-based scheduling on its own proved to severely stressful on the network due to simultaneous tariff activation, but was in combination with the ANM algorithm able to alleviate network congestion. Total charging costs were reduced by 36 percent, although 10 percent of EVs were not able to fulfill their charging requirements, indicating difficulties in societal adaptation in contrast to monetary lucrativeness in future implementations.