CIRED Workshop - Helsinki 14-15 June 2016 Paper 0405- Paper No 0405 Page 1 / 4 MULTI-AGENT CONTROL SYSTEM FOR THE EXPLOITATION OF VEHICLE TO GRID IN ACTIVE LV NETWORKS Susanna MOCCI, Nicola NATALE, Fabrizio PILO, Simona RUGGERI Department of Electrical and Electronic Engineering, University of Cagliari, Cagliari, ITALY susanna.mocci@diee.unica.it, nicola.natale@diee.unica.it, pilo@diee.unica.it, simona.ruggeri@diee.unica.it ABSTRACT Worldwide considerable resources in terms of R&D have commissioned in the field of electrical mobility. Due to the significant dissemination of Electric Vehicles (EVs) predicted in the near future, the concept of vehicle-to-grid (V2G) will be practically achievable, to improve the integration of renewable energy into distribution networks and leads to economic and environment benefits, supporting the grid operation, reducing substation transformers and line overloading to defer or even avoid investments. In the paper, an intelligent and decentralized Multi Agent System for managing EV charging/discharging in LV distribution networks is proposed. INTRODUCTION Economical, political and environmental issues increase the interest on electrical mobility. In fact, EVs can significantly reduce the usage of fossil fuels, and consequently the greenhouse gas global emission, ensuring at the same time (together with renewable energy resource, RES) to diversify the energy mix. The EVs impact on the grid depends on how and when their charging is done. The worst case would be all drivers recharging in the evening when they return from work, drawing maximum charging current, when the load demand is at its maximum. Power congestions and voltage fluctuations, overload of distribution transformers and cables, mostly in the LV distribution network, are to be expected as well as the increasing of energy losses. Different methods for controlling the EVs charging can be used in the distribution networks: centralized or decentralized approaches are both valid options, with different actors involved [1]. In [2] the authors proposed a Multi-Agent System (MAS) for demand side integration (including EVs and responsive loads) in LV networks, based on a Master-Slave interaction that allows finding a global optimum without a direct control of each resource. In the recent literature the benefits derived from the EV grid support services have been highlighted, considering unidirectional and especially bidirectional power flow between EV and the grid [3]. Bidirectional operation can maximize market and network benefits. In fact, EVs can inject both active and reactive powers and offer network services. V2G means the exploitation of vehicular energy storage systems, e.g. for mitigating the harmful intermittency nature RES. In the paper, a decentralized MAS for managing both EV charging/discharging in LV distribution networks is proposed, taking a step forward in the research activity developed in [2]. The paper shows the benefits of including the V2G concept in the optimization process, supporting the grid operation for relieving contingencies, reducing substation transformers and line overloading. Domestic EVs charging stations behave as intelligent autonomous agents with a certain level of autonomy, making decisions based on EV owner’s preferences, charging infrastructure parameters, and taking into account the state of the network and energy market conditions. The scheduling is performed in such a way that the charging program keeps the voltage within a prefixed regulation band. EV INTEGRATION In the past years, Europe, driven by economical, political and environmental issues, has gone through the initial adoption phase of electric mobility. The success of EVs depends on a variety of factors. Without smart technologies, the EV integration calls for grid reinforcements and, generally, for investments to widening the existing hosting capacity [1]. A smart alternative is the use of a charging strategy, based on time/price signals (real-time or day-ahead locational marginal prices), to move the recharge when electricity prices are lowest (e.g., nighttime charging). The intelligent charging is a step forward, in which information about electric grid conditions (voltage node, current branch), customer needs and behaviours, and EVs’ main characteristics are used to determine an optimal charging schedule. Vehicle to grid (V2G) With V2G, the vehicles’ batteries contribute to the operation of the network by providing energy when is necessary for system operation and charging when it is good for the system. The stored energy can then be delivered back to the grid during the parking hours, giving potential benefits (provision of several ancillary services like peak power and spinning reserve). The V2G concept allows an integration of RES by exploiting the batteries as an energy storage system to help facing the intermittent nature of renewable energy sources like wind and solar energy [3]-[5]. DSOs (and TSOs) have understood the potential of energy storages because of their remarkable synergy with RES, security congestion management, losses reduction, power quality, and voltage control [6]. V2G supports the grid operation, absorbing some energy in off-load hours and to supply a certain amount of power/energy during peak periods so reducing substation transformers and line