Electric Power Systems

Advanced Forecasting Techniques and Optimal Generation Scheduling

João P. S. Catalão

• Publisher: CRC Press
• Year: 2017
• ISBN:
  9781351832571
• Edition: 1
• Topic: Sciences physiques
• Subtopic: Physique

Electric Power Systems: Advanced Forecasting Techniques and Optimal Generation Scheduling helps readers develop their skills in modeling, simulating, and optimizing electric power systems. Carefully balancing theory and practice, it presents novel, cutting-edge developments in forecasting and scheduling. The focus is on understanding and solving pivotal problems in the management of electric power generation systems.

Methods for Coping with Uncertainty and Risk in Electric Power Generation

Outlining real-world problems, the book begins with an overview of electric power generation systems. Since the ability to cope with uncertainty and risk is crucial for power generating companies, the second part of the book examines the latest methods and models for self-scheduling, load forecasting, short-term electricity price forecasting, and wind power forecasting.

Toward Optimal Coordination between Hydro, Thermal, and Wind Power

Using case studies, the third part of the book investigates how to achieve the most favorable use of available energy sources. Chapters in this section discuss price-based scheduling for generating companies, optimal scheduling of a hydro producer, hydro-thermal coordination, unit commitment with wind generators, and optimal optimization of multigeneration systems.

Written in a pedagogical style that will appeal to graduate students, the book also expands on research results that are useful for engineers and researchers. It presents the latest techniques in increasingly important areas of power system operations and planning.

1

Overview of Electric Power Generation Systems

Cláudio Monteiro

1.1 Introduction

1.2 Power Generation Technologies

Generation, the Heart and Brain of the Power System • Thermal Power Generation • Hydropower Generation • Wind Power Generation • Other Nonscheduled Power Generation • Storage Technologies

1.3 Operation of Power System Generation

Power Generation Control • Scheduling • Reserve Requirements • Unit Commitment • Economic Dispatch • Hydrothermal Coordination • Scheduling with Nondispatchable Generation • Scheduling in a Market Context

1.4 Challenges of Future Power Generation

References

1.1 Introduction

Power systems are one of the largest and most complex engineering systems created by mankind. The importance of these systems is unquestionable, giving us a product that was, is, and will be the support for the development of modern society.

Historically, all power systems have developed in a similar way although with some technical variants [1]. They started their history as small isolated systems, powered by small production units with autonomous control and local distribution networks with small extension. These small systems have evolved in size and extension clustering in interconnected systems, raising the size of generation units to increasingly larger and more efficient units, using various energy sources. Large generation systems feed larger service areas, and consequently need technological solutions to transmit power through long distances, requiring transmission and subtransmission systems with multiple stages of voltage level. Over the years, power systems have continued to grow and expand to the remotest locations where electricity is needed. This evolution resulted in a very efficient generation system, interconnected by transmission systems capable of delivering energy for hundreds of miles with little energy loss, and versatile distribution systems, with high reliability and quality, safe for end users.

Thus, the traditional power system is composed by a large centralized generation system, a very high-voltage transmission, a high-voltage subtransmission system, and a distribution system. Because this power system structure is an optimized solution, this configuration has successfully persisted through decades with high efficiency and quality of service. In recent decades, however, the threat of the unsustainability of the system was evident, from the perspective of the environment and the security of the supply. The conventional energy sources threaten not to be enough to cover the increasing consumption and are not sufficiently clean to mitigate the present environmental problems of the planet. This is how a new paradigm of dispersed generation based on renewable energies emerged [2]. The purpose of this new paradigm is to collect an endogenous and clean energy resource. Because of the dispersed nature of this resource, the change of the power system at a conceptual, technological, and organizational level is inevitable. This new paradigm develops dispersed and renewable generation on a larger scale, which is intermittent and difficult to control, requiring new technical approaches to control the systems.

It is amazing how actual power systems, with their dimension and complexity, are controlled. As there is a minimum storage component in power systems, it is necessary to ensure a perfect balance between production and consumption for every millisecond. We can imagine the difficulty and complexity of control in a system where most variables are not directly controllable. Power systems have the particular characteristic of being controlled mainly from the power flow itself rather than through a separated and dedicated information system. This type of control is possible when the generating units in a system are controllable; however, in recent decades, many new components that are not controllable have emerged, especially dispersed generation. The challenge of the power systems of the future is to maintain the same quality control with less controllable variables and less direct control over the components of the system.

Not only the changes in technological paradigm but also organizational changes have brought uncertainty to power systems. The change to deregulated organizational systems, unbounded organizational structures, and market-oriented approaches originate the loss of uniqueness and centralization in the management and control system. The number of independent agents in the system (e.g., independent power producers, market agents) with the capability influencing the control is increasing. Moreover, these agents work in a competitive environment which means that information, relevant for strategic decisions, is kept secret among agents. All this additional uncertainty in the information can only result in a lower capacity to control and optimize the management of power systems.

Throughout this chapter we will provide a brief overview of electrical power systems and give a perspective of its characteristics and control variables in order to lay the foundations for understanding the issues addressed in this book. Here, we will also introduce the challenges and possible solutions for this environment in order to change the power systems.

1.2 Power Generation Technologies

1.2.1 Generation, the Heart and Brain of the Power System

Generation is certainly the heart of the power system—it is from this unit that the power flows through the whole system to reach the consumers. In contrast with other network systems, the flow is not controlled at the transmission level but is mainly controlled in the generation units. Therefore, we may say that generation is the heart and brain of energy systems.

The generation system is very diversified. It has integrated different types of energy conversion technologies: thermal, wind, solar, and hydropower. It has various types of primary energy; for instance, in the case of thermal plants, the energy source can be gas, coal, fuel oil, geothermal, biomass, or solar. The generation system is also varied with regard to the size of the generation units: the system can integrate GW units, like nuclear power plants, or small generation, like residential photovoltaic microgeneration. The generation system is also diversified in terms of the geographic distribution; big conventional units have the advantage of being centralized, but renewable resources must be collected from dispersed areas with small-sized units. The different sizes of the systems respond to the optimal economical solution but are restricted by technical, geographical, and environmental constraints. This diversification of the generation technologies is a major advantage for the power system, mixing alternative energy resources and technologies, allowing more independence from fossil energy sources and energy markets providing security of supply. On the other hand, the management of such technological diversity implies greater complexity in managing and controlling the system [3]. We will address this problem and provide solutions throughout this book.

Some of the technologies can be scheduled. For fuel-based technologies, we can manage the storage and usage of the resource, as is the case for most thermal and some hydropower generation technologies. But for other technologies, we are unable to control the resource, as is the case for wind and solar photovoltaic technologies. In these cases there are independent variables that cannot be controlled. There are two possible ways of controlling them: one is to waste the energy, not using it when it is available, or forecasting the resource availability [4]; the second is the approach followed in this book, which is obviously the most intelligent approach for optimizing the internal resources.

The seasonality of the resource is an important issue for the scope of scheduling solutions. This seasonality differs from region to region and its impact depends on the proportion of different technologies. Different energy sources have different availability throughout the year. For instance, in Europe, during the winter months, the availability of water resources is very high and sometimes excessive. On the other hand, during the summer months, the availability of solar resource is high. This complementarity is very valuable for optimizing the system, but in some cases there is seasonal coincidence of the resources. For instance, wind resource in Europe is generally higher during the winter months, overlapping with the period of high penetration of hydropower generation. This can be a problem for power systems that use reservoirs to store excess wind energy, because this seasonal period reservoir no longer has the capacity to store wind power by pumping water to high reservoirs [5].

The variability and intermittency of the resource are important challenges to be overcome in generation scheduling. Some technologies, such as solar and wind, have very fast variability. In an area of 500 × 500 km² the wind resource can vary about 10% of the maximum capacity in 1 h [6]. Generation based on solar resources can be even faster, with variations of 30%. The aggregation of a wide geographical extent of dispersed generation softens this variability, but in a different way for each technology; for instance, it softens the variability, but this effect is more important for photovoltaic power than for wind or hydropower. This is due to the differences in the correlation of production at different distances.

However, the intermittency and variability do not depend only on meteorological aspects. Sometimes the mechanisms of tariffs and market signals cause undesirable variability for generation. For instance, step tariff periods in small hydropower generation, with lower price for off-peak and higher price for non-off-peak times, cause an artificial step variation of 40% generation in less than 1 h. Signals from market prices may also cause artificial variations in scheduling generation. This is more evident in technologies with more ability to store the primary resource, such as hydropower, biomass, and cogeneration. However, usually, market price signals are a positive contribution to the system control and a mitigation of the generation variability.

The regulatory aspects and the ownership of the generation also have influence on the management of the production mix. For instance, there are some generation units that are available to be changed according to the needs of the system. There are other generation units that just follow the rules of the market and the power system operators are only able to impose certain restrictions. There are also independent power producers that, in most cases, are totally uncontrollable and system operators can only impose a few restrictions in extreme cases of system noncontrollability.

1.2.2 Thermal Power Generation

As mentioned earlier, there are various characteristics for thermal power stations. In thermal generation, we include power plants based on fossil fuels and nuclear power plants. Thermal plants based on fossil fuels are classified according to the type of fuel, which can be coal, fuel oil, or gas [7]. The principle of operation follows a sequence of energy transformation. Initially, the fuel is burned in a boiler that produces water vapor. In the second stage, steam—at different pressure levels—is transformed into mechanical energy through a steam turbine. Finally, the mechanical energy is converted into electrical energy. The efficiency of the plant depends on the calorific value of fuel, but in general is less than 45% for steam-cycle power plants. For environmental or economic reasons, in many cases thermal generation units are converted to use other types of fuel. Some plants that were originally designed for coal were later converted to oil, converted back to coal, and then converted to gas. Because of thermal inertia of steam boilers, which is usually more than 6 h, but can reach 10 h for a completely cold start, thermal power stations are restricted to the temporal switch-on and switch-off, and consequently are slow and inflexible, conditioning the strategy of scheduling (unit commitment). For this reason, usually thermal power stations operate frequently on standby, without production, keeping warm for quick starts; of course, this has an associated cost to be considered in scheduling strategy.

Due to the limits of combustion stability of the boilers, steam power plants have a technical minimum in the order of 30–40% of nominal power; it is not recommended to operate the plant below this value, as it could cause rapid decrease in efficiency. Some plants are more flexible and can operate under a stop–start daily cycle. Other plants with higher thermal inertia are slower and can only operate on a weekly cycle, stopping and starting once a week. The thermal constants of the boiler impose the speed with which the power plant can vary the generation level. Thus, each plant has its own characteristic of ramp-up or ramp-down.

In addition to steam thermal power plants, there are two other types of plants based on fossil fuels. One of these is the gas turbine power plants, in which turbines gas is burned with air under pressure and the turbine converts the high temperature and pressure into mechanical energy converted into electricity by the power generator coupled to the same axe. The other type of thermal power plant is the combined-cycle type. This type combines a closed-loop steam cycle turbine with an open-cycle gas turbine. The main cycle is the gas turbine cycle, in which a compressor, coupled to the turbine axe, absorbs, compresses, and injects air into the combustion chamber. The hot gas expands in the turbine, making the first extraction of mechanical energy. From this first stage, the resulting gas that remains at a relatively high temperature is used to produce steam and operate the steam turbine, taking full advantage of the calorific value of fuel. The combined-cycle power plant has a high efficiency of about 60%. The plant also has the advantage of flexibility—it can operate with a fast start or fast ramping similar to typical gas turbines. In terms of operation, a combined-cycle power plant can have a cold start in just 1–2 h, but if needed, this time can be just a few minutes starting as a simple gas turbine. For this reason, and for economic and environmental advantages associated with the use of natural gas as fuel, the use of the combined-cycle plant is growing. The good competition of investment and operating costs makes a combined-cycle power plant a very interesting solution for countries that have natural gas available with some security of supply.

Nuclear power plants are large-sized generation units, with about 1000 MW, in contrast with other thermal power plants with a typical size of 500 MW. There are power plants that produce at a constant pattern, because of the dangers of the variation in the operation conditions of the refrigeration system. Basically, a nuclear power plant consists of a nuclear reactor based on a fission process producing a lot of heat. This heat is extracted through a heat transfer fluid and is transferred through a heat exchanger to a steam circuit. Steam thermal energy is converted first into mechanical energy and then into electricity as in the conventional steam power plants. For reasons associated with the risk of failure of the cooling system, nuclear plants have very little flexibility: they can never stop and production can vary with very slow rates. For this reason, they have very important restrictions for scheduling optimization; it is not an easily controllable variable despite being a thermal generation. However, in market environments, scheduling is very influenced by nuclear generation, through its influence on market prices. This influence may be direct, when markets integrate nuclear generation, or indirect, when the price signals appear in the electricity import transactions with neighborhood markets and networks. During off-peak periods, the price signal influenced by nuclear generation can be very low. In fact, these prices are not the cost of generation with nuclear but the cost associated with the risk of nongeneration.

For the scheduling problem, specific characteristics for each thermal unit are required:

• Maximum generation limits (MW) correspond to the maximum overload operation of the power plant; generally, the specific fuel consumption is high for overload operation and optimization avoids these extreme operation points.

• Minimum generation limits (MW) correspond to the minimum value required to guarantee the...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• Preface
• Acknowledgments
• Chapter Abstracts
• Editor
• Contributors
• 1 Overview of Electric Power Generation Systems
• 2 Uncertainty and Risk in Generation Scheduling
• 3 Short-Term Load Forecasting
• 4 Short-Term Electricity Price Forecasting
• 5 Short-Term Wind Power Forecasting
• 6 Price-Based Scheduling for Gencos
• 7 Optimal Self-Schedule of a Hydro Producer under Uncertainty
• 8 Hydrothermal Producer Self-Scheduling
• 9 Unit Commitment and Economic Dispatch for Operations Planning of Power Systems with Significant Installed Wind Power Capacity
• 10 Operational Optimization of Multigeneration Systems
• Index