بهره وری اقتصادی از معاملات چند جانبه و هماهنگ در بازارهای برق

This paper presents economic efficiency evaluation of electricity markets operating on the basis of a coordinated multilateral trading concept. The evaluation accounts for the overall costs of power generation, network losses, and system and unit constraints. We assume a non-collusive oligopolistic competition. An iterative Cournot model is used to characterize the competitive behavior of suppliers. A supplier maximizes the profit of each of his generating units while taking rivals' generation as given. Time span is over multiple hours. This leads to a mixed integer non-linear programming problem. We use the augmented Lagrangian approach to solve iteratively for globally optimal schedules. An IEEE 24-bus, 8-supplier, and 17-customer test system is used for illustration. The results show that such a market at times of light demands exhibits little market power, and at times of large demands exhibits a great deal of market power. This contrasts with the PCMI and HHI concentration measures, which give fixed measurement values of market power. The results of two-year (730 round) market simulations show a range of deadweight efficiency loss between 0.9 and 6% compared to that of PoolCo which results in a range between 0.5 and 10% for the same test case.

Major structural changes are sweeping the electric power industry in the United States. A driving force is to increase competition and open transmission access between states. Within the states, large efforts are being made to restructure local power industries. An emerging model is being implemented in the state of California. In this model, there is an energy marketplace composed of a power exchange (PX), brokers, and an independent system operator (ISO). Undoubtedly, this new structure will continue to evolve. To gain insights into this evolution, we study two basic organizational concepts: PoolCo presented in [4] and [13] and the coordinated multilateral trading (CMT) formalized by Wu and Varaiya [14]. The concept of the CMT leads to a decentralized market. Suppliers and customers arrange their trades independently or through brokers. Brokers and the ISO coordinate with each other to meet transmission constraints, achieve economics and allocate losses. This concept was shown to achieve economic optimality under the assumption of perfect competition: competitors have no market power and are willing to participate in profitable trades. Wu et al. used a 3-bus system to show the result. If there is a degree of market power, however, economic optimality may not be achieved. Much research has shown that market power is a key issue in existing and emerging electricity markets [6], [9] and [15]. Market power can be characterized by a measure of market concentration such as the Herfindahl–Hirschman index (HHI) or by a measure of the price–cost margin index (PCMI) (often referred to as the Lerner index). The HHI measures the effective number of suppliers in a market. The PCMI, on the other hand, measures the degree to which prices exceed marginal costs. These measures were noted to be inadequate [6] and [12]. The modeling approaches, mainly, simulate markets as if elasticity of demand and market shares are known. The result of these measures is fixed and does not capture market variations. Ref. [12] presents a market power index as a ratio of the forecasted and total available supply. Meanwhile, the deadweight efficiency loss is a standard for evaluating economic efficiency. It is equal to the difference between the optimal social welfare and the social welfare due to market power. As market conditions vary, the deadweight loss varies accordingly. This paper presents an evaluation of the economic efficiency of the CMT concept. We set up a detailed market model. The model includes the overall costs of power generation, network losses, and system and unit constraints. We assume a non-collusive oligopolistic competition. An iterative Cournot model is used to characterize the competitive behavior of suppliers: a supplier maximizes the profit of each of his generating units while taking rivals' generation as given. Time span is over multiple hours. This leads to a mixed integer non-linear programming problem. The inclusion of network losses and costs has an important impact on the solution methods to be applied to this problem. Optimal scheduling of a generating unit depends on scheduling of other units. We use the augmented Lagrangian approach presented in [1] and [2] to solve iteratively for globally optimal schedules. The references mainly represent the main results of the approach and associated computational procedures for solving the hydrothermal scheduling problem. The paper is organized as follows: an overview of definitions and notations is given in Section 2. Section 3 introduces a modeling framework. Economic and PoolCo system evaluation is discussed in Section 4. Section 5 presents the results of a test case. A comparative discussion between PoolCo and CMT markets is given in Section 6. The paper is finally concluded in Section 7.

The framework described in this paper gives a quantitative evaluation of the economic efficiency of the CMT market structure. We use a detailed market model. The model accounts for the overall costs of power generation, network losses, and system and unit constraints. It offers a significant advantage over concentration analyses. The results show that a market at times of light demands exhibits little market power, and at times of large demands exhibits a great deal of market power. This contrasts with the PCMI and HHI concentration measures, which give fixed measures of market power. The results of two years (730) market simulations show a range of deadweight efficiency loss between 0.9 and 6% compared to that of PoolCo which results in a range between 0.5 and 10% for the same test cases. There are other issues which may impact economic efficiency, but are not included in this framework. Among these are: customer strategic behavior, long-term contractual agreements, and low-income customer protection. Furthermore, we have assumed that every supplier knows and takes into account rivals' generation outputs. In reality, the competitors may have more complex behavior. In fact, the role of the models developed and constrained optimization is to show the true impact of the variables related to the economic efficiency on alternative structures of the industry. Great forces such as regulation, technology and policy also drive the future industry. This work is part of an ongoing research effort to develop an integrated evaluation framework. Our future work will include the element of reliability in the evaluation process.