Electricity Markets with Increasing Levels of Renewable Generation: Structure, Operation, Agent-based Simulation, and Emerging Designs

This chapter provides a comprehensive review of four key electricity markets:

It also discusses how the outcomes of each of these markets may be impacted by the introduction of high penetrations of variable generation. Furthermore, the chapter examines considerations needed to ensure that wholesale market designs are inclusive of emerging technologies, such as demand response, distributed generation, and distributed storage.

The electric power industry has undergone a sweep restructuring resulting in the emergence of electricity markets (EMs) worldwide. The trend towards EMs has led to extensive efforts by the research community to develop optimization and equilibrium models adapted to the new competitive industry. The complexity of EMs calls, however, for richer and more flexible modeling techniques. Agent-based simulation is a relatively new approach relying on advanced social science methods as well as established engineering modeling techniques. The agent-based approach presents itself as a promising approach to accurately model and analyze in detail the behavior of EMs over time. Agent-based simulation of EMs is, at the time of writing, an active area of research and a number of prominent models and systems have been proposed in energy-related journals. These high-quality scientific contributions exhibit fairly different features and make use of a diverse range of concepts. Currently, there seems to be no agreed framework to compare the usage of specific concepts in one contribution with usage in other contributions, nor to compare disparate research efforts. This chapter and its companion (Chap. 3) claim that the development of such a framework can be an important step to provide a coherent set of concepts related to the area, to assess progress in the area, and to facilitate the development of future models and systems. Accordingly, this chapter (Part I) and Chap. 3 (Part II) introduce a generic framework for agent-based simulation of EMs. The complete framework includes three groups (or categories) of dimensions: market architecture, market structure and software agents. This chapter describes, in considerable detail, the components of the first two groups of dimensions, notably the architecture and core structure of power markets. The third and last group of dimensions is the subject of Chap. 3.

Agent technology is a relatively new and rapidly expanding area of research and development. The major motivations for the increasing interest in intelligent agents and multi-agent systems include the ability to provide solutions to problems that can naturally be regarded as a society of autonomous interacting components, to solve problems that are too large for a centralized agent to solve, and to provide solutions in situations where expertise is distributed. Electricity markets (EMs) are complex distributed systems, typically involving a variety of transactive techniques (e.g., centralized and bilateral market clearing). The agent-based approach is an ideal fit to the naturally distributed domain of EMs. Accordingly, a number of agent-based models and systems for EMs have been proposed in the technical literature. These models and systems exhibit fairly different features and make use of a diverse range of concepts. At present, there seems to be no agreed framework to analyze and compare disparate research efforts. Chapter 2 and this companion chapter claim that such a framework can be very important and instructive, helping to understand the interrelationships of disparate research efforts. Accordingly, Chap. 2 (Part I) and this chapter (Part II) introduce a generic framework for agent-based simulation of EMs. The complete framework includes three groups (or categories) of dimensions: market architecture, market structure and software agents. The first two groups were the subject of Chap. 2. This chapter discusses in considerable detail the last group of dimensions, labeled “software agents”, and composed by two distinct yet interrelated dimensions: agent architectures and agent capabilities.

Denmark has the highest proportion of wind power in the world. Wind power provided a world record of 39.1% of the total annual Danish electricity consumption in 2014 with as much as 51.7% in Western Denmark. Many would argue that the present power markets are not designed for such high shares of wind power production and that it would be hard to get good and stable prices. However, analyses in this chapter show that the Nordic power market works, extreme events have been few, and the current infrastructure and market organization has been able to handle the amount of wind power installed so far. It is found that geographical bidding areas for the wholesale electricity market reflect external transmission constraints caused by wind power. The analyses in this chapter use hourly data from West Denmark—which has the highest share of wind energy in Denmark and which is a separate price area at the Nordic power exchange. Data have been collected from the last ten years and periods with extreme wind conditions are used as case studies to illustrate the robustness of our findings.

Defining flexibility has been a challenge that a number of industry members and researchers have attempted to address in recent years. With increased variability and uncertainty of variable generation (VG), the resources on the system will have to be more flexible to adjust output, so that power output ranges, power ramp rates, and energy duration sustainability are sufficient to meet the needs of balancing supply with demand at various operational timescales. This chapter discusses whether existing market designs provide adequate incentives for resources to offer their flexibility into the market to meet the increased levels of variability and uncertainty introduced by VG in the short-term operational time frame. It presents a definition of flexibility and discusses how increased levels of VG require increased needs for flexibility on power systems. Following this introductory material, the chapter examines how existing market designs ensure that resources have the right incentives to provide increased flexibility, and then discusses a number of emerging market design elements that impact flexibility incentives.

Variable generation (VG) can reduce market prices over time and also the energy that other suppliers can sell in the market. The suppliers that are needed to provide capacity and flexibility to meet the long-term reliability requirements may, therefore, earn less revenue. This chapter discusses the topics of resource adequacy and revenue sufficiency—that is, determining and acquiring the quantity of capacity that will be needed at some future date and ensuring that those suppliers that offer the capacity receive sufficient revenue to recover their costs. The focus is on the investment time horizon and the installation of sufficient generation capability. First, the chapter discusses resource adequacy, including newer methods of determining adequacy metrics. The chapter then focuses on revenue sufficiency and how suppliers have sufficient opportunity to recover their total costs. The chapter closes with a description of the mechanisms traditionally adopted by electricity markets to mitigate the issues of resource adequacy and revenue sufficiency and discusses the most recent market design changes to address these issues.

The requirements concerning the reliability in power supply are high. This chapter addresses the issue of system adequacy, i.e., the need of enough installed capacity in each area to meet the load with an acceptable reliability. The challenge in liberalized markets is that the utilization time of rarely used peak units is so low that they will require extreme prices in order to be profitable. This has led to different methods including the creation of different types of capacity markets. The aim of this chapter is to analyze the connection between peak prices, system adequacy, needed size of a strategic reserve, i.e., the volume of the capacity market, and the impact of wind power on strategic reserves. The chapter uses data from Sweden to perform a detailed analysis of the influence of renewable generation on the capacity adequacy requirements for three different situations.

This chapter presents the key features of an agent-based simulation tool, called MATREM (for Multi-Agent TRading in Electricity Markets). The tool allows the user to conduct a wide range of simulations regarding the behavior and outcomes of electricity markets (EMs), including markets with large penetrations of renewable energy. In each simulation, different autonomous software agents are used to capture the heterogeneity of EMs, notably generating companies (GenCos), retailers (RetailCos), aggregators, consumers, market operators (MOs) and system operators (SOs). The agents are essentially computer systems capable of flexible, autonomous action and able to interact, when appropriate, with other agents to meet their design objectives. They are able to generate plans and execute actions according to a well-known practical reasoning model—the belief-desire-intention (BDI) model. MATREM supports two centralized markets (a day-ahead market and an intra-day market), a bilateral market for trading standardized future contracts (a futures market), and a marketplace for negotiating the terms and conditions of two types of tailored (or customized) long-term bilateral contracts: forward contracts and contracts for difference. The tool is currently being developed using both JADE—the JAVA Agent DEvelopment framework—and Jadex—the BDI reasoning engine that runs over JADE, enabling the development of BDI agents. A graphical interface allows the user to specify, monitor and steer all simulations. The human-computer interaction paradigm is based on a creative integration of direct manipulation interface techniques with intelligent assistant agents. The target platform for the system is a 32/64-bit computer running Microsoft Windows.

The growth of wind power generation over the past decade has surpassed all expectations. The cost of the wind energy support policy was, however, quite significant and to a large extent has led to somewhat intensive debates. The merit order effect (MOE) is an important aspect to be considered in all debates, albeit sometimes oversimplified or even ignored. Accordingly, the central goal of this chapter is to analyze and quantify the reduction in the Portuguese day-ahead market prices achieved by wind power as a result of the MOE in the first half of 2016. The results generated by an agent-based simulation tool, called MATREM, indicate a price reduction of about 17 €/MWh for the entire study period. The (total) financial volume of the MOE reached the considerable value of 391.055 million €. Especially noteworthy is the net cost of the wind energy support policy, which takes into account the feed-in tariff, the market value of the wind electricity, and the financial volume of the MOE. This cost reached the value of \(-8.248\) million € in January 2016, a negative value, indicating that a net profit has occurred in the month. The (total) net cost was 69.011 million € during the study period. Although considerable, this cost should be interpreted carefully, since it did not take into account the interaction of wind generation with the climate policy and the EU emission trading system (i.e., the carbon price effect on the electricity market).

The electricity industry is undergoing a deep transformation as Europe moves towards a greener, healthier future—the growth of renewable generation has surpassed all expectations and demand response (DR) has emerged as a key element of market design. Most European countries have already opened their markets to the participation of demand response and, over the long-term, DR will probably reach its full potential as the entire range of DR programs will be made available to retail customers. To date, however, progress has been only gradual. There is currently a need to understand and quantify the major impacts and benefits of DR, to facilitate an effective implementation of DR programs. Accordingly, this chapter investigates the impact of different levels of DR on the Iberian market prices, during the period 2014–2017, and analyzes the potential benefits for market participants and retail customers. The results generated by an agent-based simulation tool, called MATREM, are striking. In the year 2017, for instance, a modest load reduction of 5% when prices rose above 80 €/MWh yielded the (very large) benefit of 76.62 million €. Also, the same decrease in load when prices exceeded 90 €/MWh provided the (still large) benefit of 39.05 million €. The chapter concludes with specific recommendations—for consideration by state institutions, system operators, electric utilities and other market participants—to foster demand response in Portugal through both incentive-based and price-based programs.

The increasing penetration of distributed energy sources, mainly based on renewable generation, calls for an urgent emergence of novel advanced methods to deal with the associated problems. The consensus behind smart grids (SGs) as one of the most promising solutions for the massive integration of renewable energy sources in power systems has led to the development of several prototypes that aim at testing and validating SG methodologies. The urgent need to accommodate such resources require alternative solutions. This chapter presents a multi-agent based SG simulation platform connected to physical resources, so that realistic scenarios can be simulated. The SG simulator is also connected to the Multi-Agent Simulator of Competitive Electricity Markets, which provides a solid framework for the simulation of electricity markets. The cooperation between the two simulation platforms provides huge studying opportunities under different perspectives, resulting in an important contribution to the fields of transactive energy, electricity markets, and SGs. A case study is presented, showing the potentialities for interaction between players of the two ecosystems: a SG operator, which manages the internal resources of a SG, is able to participate in electricity market negotiations to trade the necessary amounts of power to fulfill the needs of SG consumers.