Energy Storage in Power Systems

Foreword

In response to the move in Europe towards a more sustainable, reliable, and cost-efficient society, European energy policy has set ambitious goals for the European electricity system, fixing the objective of at least 80% decarbonization by 2050.

Distribution networks represent 95% of the electricity grids in Europe. They are therefore a precondition for the retail markets (and also for wholesale) and for the sustainable development of cities and communities, new jobs, and growth.

The transition to a low-carbon society will boost Europe’s economy thanks to increased innovation and investment in clean technologies and low- or zero-carbon energy. A low-carbon economy implies a much greater need for renewable energy sources (RES), which are often geographically distributed (90% of the RES in the European Union (EU) are connected to the distribution networks), and also the integration of electric vehicles, which will represent a big shift in demand. Complementary IT solutions are being introduced to electricity networks at both the transmission and the distribution level, adding communication, sensors, and automation to actively manage the new and variable generation and demand. We call these Smart Grid technologies.

Distribution system operators (DSOs) and transmission system operators (TSOs) are responsible for the security of supply and the quality of service on their respective networks. It is EU policy that is driving the need for a reengineering of our electricity networks. New system challenges, including at the distribution level, lead to new network challenges for the pan-European transmission network. Hence, each DSO and TSO in the EU will have to evolve progressively from a business as usual approach to a proactive approach in order to avoid becoming a bottleneck in the future European electricity system.

It is perhaps surprising that the technologies required to address the new network challenges are, for the most part, not where the research and development (R&D) efforts are most needed. Overall, such inevitable evolution will also require the adaptation of existing regulatory regimes and business models more than technologies.

If the EU is to complete a real internal energy market, regulated companies must play a market facilitation role. TSOs, DSOs, regulators, power generators, retailers, traders, industrial consumers, and storage and RES project developers are all playing key roles in delivering an efficient electricity market. To reach the right setup, however, will involve a multidisciplinary approach to research activities, whereby network operators, manufacturers, and economists must cooperate closely in addressing the many barriers that have been identified - regulatory barriers being an important hurdle to jump.

National regulation still in operation continues to be based on the former design of electricity systems: predictable, controllable, and centralized energy generation, delivering power one-directionally through transmission and then distribution lines, with network charges calculated according to this split.

Now, more and less predictable sources of energy such as wind and solar are being generated locally and connected directly to distribution and sometimes transmission networks (larger plants). This means less controllable generation of energy, the need for bidirectional power flows and the transformation of ordinary consumers to prosumers. One of the key objectives of network operators, therefore, is to be able to use innovative approaches that are applicable in multivendor environments.

To expand on this last point, the extension and reinforcement of networks in the volume and at the rate required in the lead-up to 2020, 2030, and 2050 will be a costly endeavor. As a result, the DSO–TSO community has identified a number of grid users – demand-side response, electricity storage modules, large consumers, and even aggregated household consumer generators – as potential offerers of what are called system flexibility services, which could, in conjunction with smart technologies, reduce the need for investment in traditional assets.

It is not, however, so easy to make use of such flexibility services under the relatively new laws of the Third Energy (Liberalization) Package, which have imposed a separation of all market activities (generation and retail) and energy networks (transmission and distribution). If one considers this in the context of developing a real market for such services, the need for R&D to address the possible setups and business models becomes ever more apparent. Then there are difficult questions around the funding of R&D and demonstrations of innovative developments under the national regulatory frameworks (not at all available, in many cases).

The next step for the electricity networks R&D roadmap, under development from 2015, is the integration of R&D on storage technology applications into the existing roadmap - storage being able to offer an important form of flexibility. The European Commission is attributing an increasing amount of importance to the integration and alignment of R&D efforts, as well as to its policies in general. This is why, before influencing the calls under the EUs new R&D funding framework to 2020, Horizon 2020, the content of the roadmap is assessed alongside the R&D roadmaps for other energy sectors under the umbrella of the European Commissions EUs Strategic Energy Technology Plan (SET-Plan). The end result is the Integrated Roadmap, designed as the feed-in document for the Horizon 2020 annual work programs.

Storage is therefore becoming an unavoidable part of the power system, to ensure security of supply and as a crucial form of flexibility.

However, as indicated above, the regulatory frameworks in Europe are not adapted, in the majority of cases, for network operators – and certainly not for DSOs – to integrate storage into their networks; and this despite the considerable economic expenditure being devoted to research by these companies when allowed to do so by the national regulatory authorities.

The costs also remain a main reason for the lack of storage integration in the networks, and are still too high for a strong business case to be made at present. However, in places with a high renewable energies (RES) penetration, storage will be needed whatever the cost. Especially for DSOs, grid-optimized storage can help to address RES peak production and therefore congestion.

The question as to whether network operators will be able to own storage under strict regulation for high-risk and emergency situations, but operated by the market in all normal circumstances, is an issue that is and will continue to be the subject of interesting discussions for some time to come.

Ms. Ana Aguado Cornago

Secretary General of the European Distribution System Operators

for Smart Grids (EDSO)

Brussels, June 2015

Preface

From the outset, the electric power system has been designed to maintain a balance between generation and consumption in real time. This implies severe constraints regarding the short- and long-term operation of the system in terms of security, stability, and the sizing of the units. The current design paradigm is now challenged by the massive rollout of storage units in the power system. In recent years, the electric power system has been undergoing a transition caused by the massive introduction of intermittent renewable generation, which causes a need to incorporate advanced supervision and control features into the classical network operation. With the exponentially increasing numbers of units to be supervised and controlled, advanced computational methods combined with intelligent algorithms will enable the future Smart Grid. Energy storage has not been an initial driver that has triggered the Smart Grid, but it is now definitively a key part of the Smart Grid, not only facilitating the change of technology and design, but also the overlying business models.

The Smart Grid is somehow a starting point that is enabling the massive rollout of storage, leveraging the participation of novel players in the electricity markets who have different business objectives. One important feature of energy storage in power systems is the ability to smoothen intermittent renewable generation, both for large and small-sized operations. The massive rollout of renewables will drive the use of different (centralized or decentralized) storage solutions, which will create a sufficient market size for the storage technology and push the development of the technology.

The origin of this book can be traced back to 2009, when Francisco joined the Catalonia Institute for Energy Research (IREC) to start his doctoral thesis. Andreas and Oriol became his supervisors, and rapidly decided to focus the efforts on the utilization of energy storage technologies in wind power plants. We had gained some experience working in We had some experience working in the Centre d’Innovació Tecnològica en Convertidors Estàtics i Accionaments, Technical University of Catalonia (CITCEA–UPC) and IREC on electrical systems and on grid integration of wind farms in some projects with Ecotecnia (which was acquired by Alstom, becoming the wind division of the Alstom group). At that time, we started to move away from the concept of the wind farm to the more appropriate term wind power plant. Wind power was no longer a fancy green alternative source of energy, which could generate power when the wind blew. It was now part of a massive business, which already bore a very serious level of responsibility in the operation of the whole power system. Transmission system operators were drafting very demanding grid codes, in which wind farms were treated as dependable power plants.

We remember having discussions with some engineers in Ecotecnia (Alstom Renewables, wind division) about the possibility of incorporating energy storage in the wind turbines in order to provide ancillary services. Additionally, these devices could be used for other purposes, as power smoothing, correction of production forecasts, and energy market operations. While the potential of energy storage was evident, there were differing opinions on where to locate the storage, what technology to use, and how to size such energy storage systems. Some engineers supported the idea of wind turbines equipped with energy storage devices that could allow the smooth provision of power adjusted to the forecasted production levels and that could eventually provide ancillary services to the grid. Others argued that it made more sense to operate a single, larger energy storage device at the wind power plant level and provide the same services in an aggregated manner. Other colleagues stressed that eventually energy storage should be deployed on the demand side, close to the consumer, and that it should be combined with demand-side management. Finally, other engineers defended the idea that the optimal solution was to locate the energy storage devices in the distribution substations.

During the realization of the doctoral thesis, some contributions were made on the modeling and control of energy storage systems, especially flywheels combined with wind power plants. Francisco built a scaled test rig with which he could gain some practical experience and demonstrate the possibility of power smoothing using a flywheel. We also realized that there were some impressive advances in the development of energy storage technologies and also on different applications in electric power systems. For example, energy storage was being considered as the only possible solution for preventing rapid power drops in large photovoltaic power plants and in renewable power plants in general. Energy storage was also the backbone of the microgrid concept (which is absolutely necessary to balance power flows) and the lung of the Smart Grid of the future.

By the time the thesis came to an end and was successfully defended in September 2013, we realized that we were starting to understand the potential of energy storage in power systems with a high penetration of renewable energy. Our beliefs regarding the huge potential of energy storage utilization in future power systems triggered the idea of expanding the work done in the doctoral thesis, and in other projects that we had been developing, and start the adventure of writing a book on the topic. At that time, we probably did not appreciate the massive amount of work that was awaiting us when we began the preparation of this book in April 2014.

Let us move forward to spring 2015, at which time we were working to submit the manuscript to the publisher on time. We were writing this preface in the hope and belief that this book could provide some useful guidance to engineers and professionals interested in the utilization of energy storage in power systems that are rich in renewable energy sources. Nowadays, we often hear news stories about paradigm shifts and energy revolutions that will eventually change the way in which we understand electric power in our society. In all these communications, energy storage is part of the equation. We are not certain how future electrical energy systems will be shaped, but we trust that energy storage will play an important role.

According to the scope of the book, its contents are divided into eight main chapters. Chapter 1 first introduces readers to modern power systems. Electric power systems are experiencing a dramatic transformation from the conventional vertically integrated approach with few control actors, towards a system with a high penetration of renewable (and intermittent) generation and, as a consequence, a highly controlled system at any voltage level. As previously noted, such a transformation suggests the introduction of the term Smart Grid, and this is one of the main concepts underpinning future power system architectures. The Smart Grid architecture is defined in terms of domains, zones, and layers, and these are presented in the chapter. After the presentation of the power system architecture, the chapter continues with the presentation of energy management systems and the fundamentals of power system analysis. In this regard, basic concepts on optimization methods and optimal power flow computational techniques are presented. Viewed together, this results in a didactic approach to an understanding of the fundamentals of power systems. Moreover, though, the chapter also includes a practical example on load-flow calculation.

One of the main drivers of power system transformation is the field of renewable generation, and as such this is presented in Chapter 2. The chapter first discusses the contribution of the various forms of renewable energy in the worldwide energy mix. After this presentation, the chapter classifies the renewable power generation technologies into those based on rotative electrical generators, mechanically coupled to turbines or similar devices (e.g., wind turbines and hydropower); and those based on static power generation sources, producing electricity without any moving devices (e.g., photovoltaics). With regard to the former, the chapter describes wind turbine topologies in detail, and offers two numerical examples on the calculation of the power generated by both fixed- and variable-speed turbines. Finally, with regard to static renewable-based generating technologies, the chapter introduces the concept of photovoltaic generation and proposes a calculation on the analysis of PV panels. The chapter concludes with a brief presentation of the grid code requirements for the grid connection of renewables.

With the stepwise displacement of conventional generating plants by nonsynchronized renewable-based ones, the net level of synchronous power reserves in the system becomes reduced, and this can affect the frequency control in the system. For such reasons, and according to some European grid codes, wind power plants are required to provide power reserves in the same way as conventional generating units. As a contribution to the description of the requirements for the grid connection of renewables, Chapter 3 presents an extensive literature review on the European grid codes with regard to frequency support. While the chapter looks specifically at wind power plants, the results can be exported to other renewable energy generation technologies. Apart from discussing on grid codes, the chapter includes an extensive literature review on control methods for operating wind turbines, so that they can maintain a predetermined level of power reserves, thus enabling them to participate in tasks related to frequency control.

The three chapters described above serve as a good introduction to electric power systems and renewable generation. These subjects are quite pertinent, and even somehow unavoidable, for a proper understanding of the concepts presented in the rest of the book, which are all centered around energy storage technologies in power systems.

The first chapter on energy storage is Chapter 4. This chapter offers a review of the energy storage technologies that can be potentially included in the electric power system. The chapter covers a great number of technologies, such as pumped hydroelectric storage, compressed air and hydrogen-based systems, secondary batteries, flow batteries, flywheels, superconducting magnetic storage, supercapacitors, and even (although tangentially) the field of thermal storage and the power-to-gas concept. For each technology, the description includes the operating principles, the main components, and the most relevant technical characteristics. The chapter emphasizes the main differences amongst the technologies in a comprehensive manner, including some tables and graphics based on the data collected from several publications and from manufacturers’ datasheets. The final part of the chapter discusses power conversion systems for grid connection and the control of storage not synchronized with the network.

Following the description of the technology in Chapter 4, the book tackles the formulation of cost models for the economic assessment of storage technologies. A cost model considering capital, operation and maintenance, replacement, and also end-of-life costs is introduced, based on the literature. The model is demonstrated by means of a numerical example. In this example, the life-cycle costs of different storage systems – both in themselves and while providing various services in the power system – are calculated and evaluated.

The study of the inclusion of storage technologies in the power system usually requires the development of simulation platforms to virtually validate various concepts centered on the design and operation of the technology prior to the commissioning of the system. Accordingly, Chapter 6 presents averaged dynamic models, based on electrical equations, for different storage technologies such as batteries, supercapacitors, and flywheels, as well as for their corresponding power conversion systems. Ultimately, the contents of this chapter can be adopted as a practical approach to the modeling of storage systems. To demonstrate the correctness of the models and of the corresponding control algorithms for the power conversion systems to which the storage containers are attached, the chapter includes various numerical examples. These examples plot the behavior of the storage systems modeled in charge and discharge processes.

In this way, Chapters 4–6 describe the basis for storage technologies and/or offer tools for studies related to the application of the technology. The last two chapters, Chapters 7 and 8, deal specifically with the applications that energy storage systems could potentially provide in the electric power system. Since the power systems of the future will surely be characterized by increasing penetration rates of renewables, most of the storage applications discussed in these chapters are closely related to renewable generation. Chapter 7 presents the potential for short-term applications; that is, for those applications requiring storage in order to rapidly inject or absorb power, over short periods of time, for different purposes. Conversely, Chapter 8 refers to potential mid- and long-term applications: that is, those applications requiring the storage systems to continuously exchange power with the network over periods of hours or even days, for balancing and generation time-shifting purposes.

Both chapters include a numerical example, thus contributing to the practical scope of the book. With regard to short-term applications, a specific example on wind power smoothing with flywheels is offered. This example includes the formulation and solution of an optimization problem, which determines the theoretical optimal operation of the flywheel while providing this service. From the results of this optimization problem, a control algorithm for the flywheel to be executed in real time is derived and also validated using laboratory-scale equipment. Ultimately, the proposed exercise is a good example of the combination of different analytical tools; that is, modeling, optimization, and experimental validation. Finally, the example in Chapter 8 proposes the sizing of a battery bank and its attached power conversion system, building up an isolated power system with PV generation.

For us, writing this book has required tremendous personal effort, but it would not have been possible without the invaluable support received from a number of colleagues, in various forms. We would first like to acknowledge, with thanks, the support received from our colleagues at IREC and CITCEA–UPC: this work is the product of our professional activity over recent years, and throughout this time we have gained experience and knowledge from all of them.

Particularly related to the book, we thank Cristina Corchero and Joana Aina Ortiz for providing us with data for simulations. We thank Jordi Pegueroles and Fernando Bianchi for the design of control algorithms; José Luís Domínguez, Mikel de Prada, and Eduardo Prieto for the figures in Chapter 2; and Gerard del Rosario and Ramón Gumara for the information on laboratory equipment.

In addition, we would like to thank Ms Ana Aguado Cornago for writing the foreword to this book.

Finally, we are also grateful for the permissions received from many authors, institutions, and companies to use figures in the book. In particular, we like to acknowledge the permissions received from IRENA, Redflow Limited, Beacon Power, the World Energy Council, and Knut Erik Nielsen.

We hope that the book will prove to be useful for researchers and engineers. Comments from, and discussions with, readers with diverse backgrounds will be highly appreciated.

Francisco, Andreas, and Oriol

Barcelona, June 2015

1

An Introduction to Modern Power Systems

1.1 Introduction

Power systems are complex structures composed of an enormous number of different installations, economic actors, and – in smaller numbers – system operators. In the traditional approach, the system is dominated by economies of scale. This means that for steadily increasing consumption, a large power generation capacity is installed, mainly nuclear, coal- or gas-fired thermal, and hydroelectric. In order to guarantee the reliability of such a system, a meshed transmission grid at high voltage has to be installed, into which the generators feed. Underlying this transmission system, function of the distribution grid is to conduct the power flow at lower voltage levels to customers, at medium or low voltage. The described power flow is mainly unidirectional, from the generators to the customers, who are connected at medium or low voltage. Only a few customers are connected at high voltage, due to their high loads. Such a system is easy to control, as most of the players (the customers) are passive, only a few actors (generators and system operators) are needed to centrally control the system, and the interfaces are well defined. The most extended economic model in this context is the vertically integrated utility. However, some of the deep fundamentals on which this structure is based can be envisioned, moving from these vertically integrated utilities to the Smart Grid distribution system [1]:

Economies of scale are no longer applicable to the power system generation, due to the dramatic growth of distributed generation.

The costs of the various renewable energy technologies have declined steadily due to technological advances.

Increased environmental concerns on the part of customers and legislators.

Regulation is enabling the emergence of different players on the electricity market (retailers, energy service providers, etc.)

These fundamental changes are causing a shift from the vertically integrated approach with few control actors towards a system with a high penetration of renewable (and intermittent) generation and, as a consequence, a system that needs to be highly controlled at all voltage levels. The increasing use of renewable energy not only helps to alleviate fuel poverty, but also promotes decentralized power generation, thereby reducing the dependence on conventional grid-based energy sources. It provides electricity from small-scale generation and microgeneration; working towards reducing the increasing electricity consumption and supplying any surplus generation to the grid. Therefore, microgeneration is a key power generation trend for smart communities, both rural and urban. Distributed generation from micro–combined heat and power (CHP) installations and renewables such as small-scale wind turbines and solar photovoltaics (PV) plays a strong role in this ecosystem. New generation units from renewable energy sources must be established; however, as a result of stochastic generation, those energy resources are intermittent, and possible output fluctuations have to be balanced [2]. Energy storage applications will be used to cope with this problem [3]. All of this leads to the approach to make the grid intelligent: the Smart Grid. A Smart Grid is an electricity network that can intelligently integrate the actions of all of the users connected to it – generators, consumers, and those that do both – in order to efficiently deliver sustainable, economic, and secure electricity supplies [4]. A Smart Grid uses sensing, embedded processing, and digital communications to enable the electricity grid to be observable (able to be measured and visualized), controllable (able to be manipulated and optimized), automated (able to be adapted and to self-heal), and fully integrated (fully interoperable with existing systems, and with the capacity to incorporate a diverse set of energy sources) [5].

One prominent set of actors in modern power systems are prosumers (proactive consumers). Prosumers are common consumers who become active to help to personally improve or design the goods and services available in the marketplace, transforming both it and their own role as consumers [6]. The strategic integration of prosumers into the electricity system is a challenge. As prosumers are acting outside the boundaries of the traditional electricity companies, ordinary approaches to regulating their behavior have proved to be insufficient. The aggregated potential of flexibility makes the role of the prosumer important for energy systems with high and increasing shares of fluctuating renewable energy sources. To involve different prosumer segments, both utilities and policy need to develop novel strategies. The benefits for prosumers in modern power systems can be summarized as follows:

Economic. The Smart Grid offers the possibility of involving customers, their flexibility being used as an instrument to shed loads and secure stability. It is assumed that customers will allow the distribution system operator (DSO) access to their home automation systems, and that a value chain that links households with the transmission system operator (TSO) via the DSO will be created in such a way that the flexibility can be used systematically, as can the compensation flowing in the other direction.

Incentives. Incentives may attract customers into a demand–response regime and into distributed energy resources (DER) programs without the need for a proper compensation structure. Poor quality of supply can also be a trigger, especially when there is only one utility operating. Local DER solutions are thus a good option, although the levelized energy costs could be much higher than the supply costs from a centralized utility. Other incentives, such as environmental and social sustainability concerns, comfort, convenience, and so on, could also be drivers.

Technical. Energy storage for electricity is the main key to assuring the stability of a system with intermittent generation, at least for short periods. Ownership models and options for placement in the grid will drive very different solutions. It will be possible for electric cars to supply to the grid (vehicle to grid), which will add to the additional power system storage capability. As long as the distribution operator is in control of, or owns, these facilities, they will be operated in a different manner than if the storage is owned and operated by the community or by a third party working partly on their behalf.

The community. With DER and Smart Grid technologies, communities will gain substantial market power. Traditionally, the utility was in charge of upgrading the infrastructure in order to cater for a sufficient supply capacity and to assure quality. To build a community solution for local supply by means of Smart Grid technologies and DER seems to be the solution for future expansion, at least in rural areas.

Market and trading. New local markets and trading will arise, based on real-time trading, in order to balance the system. The flexibility of customers, local generators, and storage systems will create value on the market to balance the intermittency of renewable generation.

Social. A new form