ENERGY MARKET SIMULATION WITH TRANSMISSION SYSTEM REPRESENTATION USING GAME THEORY AND MULTI-AGENTS SYSTEMS PAULO SÉRGIO DA SILVA, LEONARDO NEPOMUCENO São Paulo State University – UNESP Bauru, Brazil E-mails: pss@eb.unesp.br , leo@feb.unesp.br Abstract This paper proposes a multi-agent system to simulate energy markets using game theory, keeping as a main focus to provide the agents with a tool to maximize their profits. One of the central features of the proposed market simulation model is its ability to represent the transmission system. This feature provided the means to evaluate some market power situations related to congestion in the grid. In this work, the proposed model is evaluated through case studies involving the pool market only. The results obtained allow the identification of market power in the systems studied, and show that the transmission system represen- tation is crucial if one is interested in more representative pool market simulation. Keywords Energy market simulation, game theory, multi-agents systems, transmission congestion. Resumo Este artigo propõem um sistema multi-agente para simular o mercado de energia usando a teoria de jogos, tendo co- mo foco principal fornecer aos agentes uma ferramenta para maximizar os seus lucros. Uma das características centrais do mo- delo de simulação de mercado proposto é a sua capacidade de representar o sistema de transmissão. Essa característica provê os meios necessários para avaliar algumas situações de poder de mercado relacionadas ao congestionamento da malha de transmis- são. Neste trabalho, o modelo proposto é avaliado por meio de estudos de casos envolvendo somente o mercado pool. Os resul- tados obtidos permitem identificar o poder de mercado nos sistemas estudados e mostram que a representação do sistema de transmissão é crucial para uma simulação mais representativa do mercado pool. Palavras-chave  Simulação de Mercados de Energia, Teoria de Jogos, Sistemas Multi-Agentes, Congestionamento da Trans- missão. 1 Introduction The energy market is going through deep restructur- ing aiming at promoting competition in generation and energy trading. In the old vertically integrated model, planning and operation were carried out by centralized optimization procedures. In the actual market structure these centralized decisions have been replaced by market-driven procedures. In such a context the agents involved in the market are faced with the problem of how to build their optimal bid- ding strategy. Game-theory-based methods are widely (Chat- topadhyay, 2004; Ferrero, Shahidehpour, Ramesh, 1997; Ferrero, Rivera, Shahidehpour, 1998; Stoft, 1999; Krishna, Ramesh, 1998; Ferreira, 2002) used to model strategic interactions between generators and thus search the market equilibrium. In general, non-cooperative strategic games and Nash equilib- rium are used, where all agents involved seek to maximize their own profits. In an energy market, network constraints induce market separation (Liu, Wu, 2007), whereby some generators can exercise its market power more easily. Thus, the study for the impacts of network constraints on the market is demanding, although most previous works describing energy market simulation have ignored them. This paper is concerned with the impact of net- work constraints in energy market simulation. This issue has been addressed in (Liu, Wu, 2007; Peng, S.; Tomsovic, 2003; Cunningham, Baldick, Baughman, 2002), with important theoretical contribution. The work here proposed extends a previously described multi-agent system and game theory (Krishna, Ramesh, 1998) for market simulation by incorporat- ing detailed transmission system representation. The games are solved using simple maxmin criterion. The simulator is tested in small test systems and the re- sults focus on the comparison between the results obtained by the market simulation method and those obtained by old vertically integrated procedures. The analysis verifies that the network congestion is of major concern when simulating energy markets. So- me typical situations have been verified where net- work congestion produces clear market power situa- tions. 2 Generation Dispatch and Market Structure In this work the market simulation is based on the following principles: • Market participants (only generators are con- sidered) submit their next-day generation bids to an independent system operator (ISO). • With submitted bids, the ISO clears the mar- ket to schedule generation and determine the market clearing price (MCP). • All agents are paid at MCP value. The agents evaluate their revenues, expenditures and final profits.