Agent-coordinated Virtual Power Plants of Electric Vehicles (Extended Abstract) Micha Kahlen Erasmus University Rotterdam Burgemeester Oudlaan 50 3062 PA Rotterdam, The Netherlands kahlen@rsm.nl Wolfgang Ketter Erasmus University Rotterdam Burgemeester Oudlaan 50 3062 PA Rotterdam, The Netherlands wketter@rsm.nl Jan van Dalen Erasmus University Rotterdam Burgemeester Oudlaan 50 3062 PA Rotterdam, The Netherlands jdalen@rsm.nl ABSTRACT Challenges with energy provision due to intermittent re- newable energy sources can be addressed with information systems in smart energy markets. One speciﬁc solution is virtual power plants (VPP) of electric vehicles (EV). The operation of the VPP is agent controlled, so that cars are charged when market prices on the wholesale market signal excess capacity and turn into a power plant when market prices signal a need for additional energy supply. We show that due to a low utilization of EVs, its storage in idle status can be used by owners of large EV ﬂeets to trade energy on the electricity wholesale market. We scrutinize and evaluate a trading strategy under diﬀerent scenarios of battery cost developments for large EV ﬂeets owners in the simulation platform Power TAC against the triple bottom line (people, planet, proﬁt). Findings indicate that people pay lower elec- tricity prices under widespread adoption of VPPs of EVs. In addition to that results show a decrease in CO2 emissions for the planet. Finally proﬁts for ﬂeet owners of EV parks are boosted, which decreases with growing market adoption. Categories and Subject Descriptors H.4.2 [Information Systems Applications]: Decision Sup- port Systems General Terms Economics; Performance; Human Factors; Measurement Keywords Innovative Applications; Smart Grid; Agents; Electric Vehi- cles; Virtual Power Plant 1. INTRODUCTION A growing support for sustainability has set in motion a move towards renewable energy sources and ultimately a low carbon economy. The EU for instance wants to increase its share of renewable energy sources to 20% by 2020. This, Appears in: Alessio Lomuscio, Paul Scerri, Ana Bazzan, and Michael Huhns (eds.), Proceedings of the 13th Inter- national Conference on Autonomous Agents and Multiagent Systems (AAMAS 2014), May 5-9, 2014, Paris, France. Copyright c  2014, International Foundation for Autonomous Agents and Multiagent Systems (www.ifaamas.org). All rights reserved. however, also bears challenges that need to be addressed for a smooth transition. The inability to predict and alter the output of solar and windpower plants under the absence of storage leads to imbalances in the generation and consump- tion of electricity destabilizing the electric grid. Smart grids are an approach to solve this problem with the help of in- formation systems and computing power. Previous research has studied the eﬀect of static battery storage on future en- ergy systems. We extend this research with an analysis of the storage capacity of Electric Vehicles (EV) that is shared for driving and balancing. Our contribution lies in the un- certainty of availability of storage by including driving pro- ﬁles. In this respect we study a large ﬂeet owner of EV that has the critical mass to participate in wholesale trad- ing with the help of the Power TAC simulation. This ﬂeet owner leases the cars to consumers and when the cars are idle he uses them as virtual power plants that generate and consume electricity in accordance with market requirements. The ﬂeet owner exploits and beneﬁts from an electricity ar- bitrage opportunity that could not previously be exploited due to unavailability of storage, which has a positive eﬀect on the triple bottom line: lower electricity prices for peo- ple, lower emissions for the planet, and proﬁts for the ﬂeet owner. 2. PROSUMER BUSINESS MODEL EV’s unique properties with large storage capacity that is quickly scalable make them particularly interesting for smart grid storage solutions. Current research focuses on smart charging that alleviates the strain that a large amount of uncoordinated charging EV have on the electric grid [5] and have addressed the concept of Vehicle-2-Grid (V2G) by making energy from the EV available to the grid [4], [6]. This way consumers not only use energy for charging but also produce energy for the grid and become prosumers. In this research we investigate how large ﬂeet owners of EV can aggregate this storage capacity in form of VPPs to actively trade energy in the electricity wholesale market. To ensure that the driving needs of individual EV drivers are not com- promised we employ intelligent software agents. These soft- ware agents combine information on driving patterns with price signals from the electricity wholesale market and based on that make decision to charge, or discharge EV in the ﬂeet as agent assisted decision support [1]. The ﬂeet owner makes additional arbitrage proﬁts by buying energy at low prices 1547