Market Power Analysis Of The Generation Firms In The Transmission Constrained Electricity Markets

For the specific features of the electricity industry, the present electricity markets may be better described in terms oligopoly than of perfect competition from which they may be rather far. In an oligopoly, as the electricity market is, the producer as the market player show a strategic behavior in which the offer submitted is different from the marginal cost and is aimed to the surplus maximization of each player. The oligopoly market clearing price, market clearing quantity and the market efficiency are depends on the strategic interactions among producers. In addition to the traditional causes of market power, in the electricity markets, the network constraints may give additional possibility of market power arising that are very specific of this contest. The aim of this thesis is to analysis the market power behaviors of the electricity producers in the competitive electricity market with the consideration of the network transmission congestion.The concentration indexes can quantify the market power measuring the capacity concentration that is one of the causes for market power. However, traditional reliance on concentration measures is likely to be inadequate for the tasks due to the specific features of the electricity industry. The negative effects, such as the low market efficiency and the high market clearing price and the low market clearing quantity, induced by the strategic behavior of the electricity producers should be simulated resorting to the suitable mathematic models. In this respect, the quantitative market power indices derived from oligopoly simulations are assumed as the ground for the market regulator to take reasonable and relevant approaches to monitor and suppress the potential market power behaviors.The mathematical model used to represent the strategic bidding behavior of electricity producer is a bi-level programming problem in which the maximization of the producer surplus, the upper optimization problem, should take into account the market clearing dispatch, the lower optimization problem. However, it is very difficult to solve the bi-level problem if the multiple trading hours of the market performance is considered. The multi agent system using the heuristic approaches should be another option to model the the complex strategic interactions among the electricity producers. Multi agent system approches do not require strict mathematic model for solving the producer surplus maximization problem but resort to the learning process of the autonomous intelligent agents, which greatly contributes to analysis the long-term oligopoly market performance.The main contributions and innovations of this thesis are listed as the follows:1) The capacity share indices, such as HHI index, are generally used to assess the market power of the electricity producers. However, such concentration indices are likely inadequate for the tasks due to the specific features of the electricity markets. This thesis proposed a batch of market power indices basing on the equilibrium oligopoly models to capture the intuitiveness of market power behaviors of the producers in the competitive electricity markets in which the traded power amount must take into account the system operational constraints.2) Current research literature focusing on the game behaviors of the electricity producers either disregard the network constraints, or simulate the transmission constrained market within a given model. Basing on the uniform test netowrk system, this thesis is targeted to analysis the impacts of the market power behaviors on different oligopoly markets with the consideration of the network constraints. The best response approach and the iterative search algorithm are employed to solve the Monopoly, Cournot, Stackelberg and Supply Function Equilibrium models for the market power analysis. 3) The transmission loss of power in the network and the game behaviors among producers are co-existed in a real electricity market. By integrating the power flow loss model and strategic bidding behavior of the producers, this thesis analyses the performance of the noligopoly electricity market with the consideration of the power flow loss along the transmission lines.4) For the non-convexity and non-linear bi-level optimization problem of the producer surplus maximization model, current literature seldom discusses the solution techniques and tricks. For different game models, this thesis developed two approaches for solving such optimization problems. One is to transfer the non-convexity and non-linear bi-level problem into linear programming by list all possible line flow states; the other is to find a good start point by using the heuristic methods, and then to solve the producer surplus maximization model.5) Currently the strategic bidding behaviors in the long-term trading period are difficult to be mathematically modeled. This thesis developed a model basing on the Watkins's Q(λ) Reinforcement Learning to simulate the long term strategic bidding behaviors of the electricity producers. Such multi agents system model can be used to define the optimal bidding strategy for each producer and, as well, to find the market equilibrium and assessing the long term market performance. 6) In general, using the deterministic approaches, the bi-level optimization problem is trapped with local solutions for its nonlinear property and is sensitive to the selection of the start points. Heuristic approaches should be another option to solve such bi-level optimization problem. For its simplicity and immune to the local optima, particle swarm optimization (PSO) is presented in this paper to find the optimal supply functions for electricity producers.7) The strategic bidding behavior of the producers results in inefficient market performence. Such negative effect is more remarkable with low demand elasticity of the electricity consumers. This thesis proposed a model to assess the role of demand elasticity in mitigating the negative effects of strategic bidding behavior of the electricity producers. The Conjecture Supply Function model is used to incorporate the exogenous changes in demand elasticity in a given market.Based on the specific technique and eonomic features of power systems, this thesis is aimed at the analysis of the performance of the oligoply electricity markets, providing the market regulator with efficient tools to monitor and mitigate the potential market power of the electricity producers.