PhD Dissertation:Danial Esmaeili Aliabadi

PhD Dissertation:Danial Esmaeili Aliabadi

LISTEN

Agent-Based Analysis of Electricity Markets

 

 

Danial Esmaeili Aliabadi
Industrial Engineering, PhD Dissertation, 2016

 

Thesis Jury

Assoc. Prof. Dr.  Güvenç Şahin (Thesis Advisor), Assoc. Prof. Dr.  Murat Kaya (Thesis Co-advisor),  Prof. Dr. Ali Rana Atılgan, Asst. Prof. Dr. Emre Çelebi, Assoc. Prof. Dr. Kemal Kılıç, Assoc. Prof. Dr. Erhun Kundakçıoğlu

 

 

Date & Time: 13th, December, 2016 – 14:40-16:30 PM

Place: FENS L027

Keywords : Deregulated electricity market, Power generation company, Collusion, Strategic bidding behavior

 

Abstract

 

As a result of liberalization, deregulated electricity markets were formed to provide affordable electricity for consumers through promoting competition. Although the new market is expected to serve this purpose, even the earliest deregulated electricity markets are prone to threats that may disrupt the competition. While the independent system operator, responsible for administering the electricity markets, aims to provide the consumer with the lowest possible electricity price, lack of competition may increase prices. We consider the effect of three major factors hand-in-hand on that may affect the level of competition in the market: the independent system operator's market-clearing mechanism as a strategic choice, strategic bidding behavior of generation companies and the transmission network.

 

We use both a mathematical modeling approach and an agent-based simulation model with a game-theoretic understanding of the market clearance mechanism involving the independent system operator and the power generation companies (as the players). The literature on deregulated electricity markets mostly focus on analyzing the behavior of power generating companies without considering the transmission network and its impact on the players’ behavior since including transmission network makes game-theoretic approaches intractable. While we consider the transmission grid in both modeling approaches, we confine the boundary of our analysis to the day-ahead market.

 

The game-theoretic understanding assists in characterizing a set of sufficient conditions for the generators to engage in a collusive behavior. These conditions are embedded into a bi-level optimization problem where the objectives of the independent systems operator are conflicting with those of the generators. We develop an algorithm to solve the multi-objective bi-level problem and we show that the generators’ optimal behavior are collusive when sufficient conditions exist.

 

 

We investigate the strategic behavior of power generation companies under different market-clearing mechanisms by an agent-based simulation model. We observe that both pricing rules and dispatch (rationing) policies can alter the behavior of generation companies. We find that pay-as-bid pricing rule together with random dispatch policy improves social welfare more than uniform pricing with equal dispatch policy. Finally, we investigate the effects of risk attitude and capacity withholding. We present a complete set of results of simulation experiments using various cases for a wide range of learning model parameters.