Local Electricity Markets

Local Electricity Markets introduces the fundamental characteristics, needs, and constraints shaping the design and implementation of local electricity markets. It addresses current proposed local market models and lessons from their limited practical implementation. The work discusses relevant decision and informatics tools considered important in the implementation of local electricity markets. It also includes a review on management and trading platforms, including commercially available tools. Aspects of local electricity market infrastructure are identified and discussed, including physical and software infrastructure. It discusses the current regulatory frameworks available for local electricity market development internationally. The work concludes with a discussion of barriers and opportunities for local electricity markets in the future.

• Delineates key components shaping the design and implementation of local electricity market structure
• Provides a coherent view on the enabling infrastructures and technologies that underpin local market expansion
• Explores the current regulatory environment for local electricity markets drawn from a global panel of contributors
• Exposes future paths toward widespread implementation of local electricity markets using an empirical review of barriers and opportunities
• Reviews relevant local electricity market case studies, pilots and demonstrators already deployed and under implementation

• Language: English
• Publisher: Academic Press
• Release date: Jul 3, 2021
• ISBN: 9780128226667

Book preview

I

Distributed energy resources as enablers of local electricity markets

Outline

Chapter 1 New electricity markets. The challenges of variable renewable energy

Chapter 2 Integration of electric vehicles in local energy markets

Chapter 3 From wholesale energy markets to local flexibility markets: structure, models and operation

Chapter 4 From the smart grid to the local electricity market

Chapter 1

New electricity markets. The challenges of variable renewable energy

Ana Estanqueiro and António Couto,    National Energy and Geology Laboratory (LNEG), Renewable Energy and Energy Efficiency Unit, Lisboa, Portugal

Abstract

The development and large-scale dissemination of the new and variable renewable technologies took place from 1990 onwards in most developed countries, in a process led by Europe. To promote the renewable sector development financial incentives, both for investment and for the payment of renewable energy, were always present. These incentives usually consisted of guaranteed feed-in tariffs that ensured a return of the investments made in this new business—thus minimizing the financial risks and building a more attractive business for private companies in the renewables sector. That approach was the main basis that essentially supported the remarkable growth of the renewable sector in Europe in the past 30 years.

Nowadays, the renewable energy sector is already mature in most aspects. The cost of generating electricity from wind or solar (photovoltaic) resources is competitive with conventional gas or coal-based technologies. However, some challenges still exist in the transition of the electrical power sector to a desirable carbon-free, near 100% renewable-based sector—and one of those main challenges is the negotiation of the electricity generated by these novel technologies, due to the time and spatial variability of the primary resources as well as their poor predictability and dispatchability of the power generated.

This chapter addresses those challenges as well as the approaches available to overcome them within competitive electricity markets.

Keywords

Renewable energy; electricity markets; power generation variability; variable renewable energy

1.1 Introduction

The development and large-scale dissemination of new renewable technologies (as wind and solar power)—referred to as variable renewable energy (VRE)—took place from 1990 onwards in most of the developed countries, in a process led by Europe. This development was anchored on the offer of financial incentives, be it for investment, or for the payment of renewable energy [1]. Those incentives were commonly called FIT—feed-in tariffs, guaranteed tariffs, or green tariffs, and consisted of ensuring investors a return profit guaranteeing the recovery of investments made, thus minimizing the financial risks and contributing to reducing the cost of the investments in the renewable energy sector [2]. That was the key basis that essentially supported the remarkable growth of the renewable sector in Europe in the past 30 years.

The majority of the European countries resorted to FIT systems that all presented similar characteristics: (1) the tariff values paid were (well) above the average energy value in the electricity markets; (2) this generation was exempted (in most countries) of costs from network usage fees or other production costs normally charged on conventional generation; and (3) in some countries the interconnection to the preexisting network was provided at low (or no) cost. Some variations exist in those incentives for generation and renewable technologies; for example, (1) Denmark followed a different, but very effective path by defining minimum percentages of renewable penetration in the energy sold by electric utilities; (2) Spain favored a mixed system, that is, a (pseudo) market share associated with the payment of a (fixed) premium above the spot market value for new renewables, as wind energy; and, outside Europe, (3) the United States has favored a (complex) system of tax incentives. In all countries, the base legislation of FITs usually defined a limited duration of application: in France, it had a maximum duration of 9 years, in Portugal initially it had a duration of 15 years and, later, it was indexed to the level of production of the installation that can vary between 9 and 15 years (depending on the period of installation and the technology) and other countries had similar rules [3]. The essential thing to keep in mind is that in all countries it was assumed that the existence of FIT was transitory and that FITs assumed an incentive role for the European energy and industrial energy sector only during the (possibly average) period of recovery of investment, guaranteeing predictability of risk that attracted private investors, by lowering financing costs.

Regardless of the natural diversity of energy policies in the various Member States, Europe is currently phasing-out FITs. In Portugal the process has already started and it is expected that in 2023 a significant part of wind generation has to move to a market regime; France and Germany have ended most of those green contracts; and in Spain FITs were suspended during the 2010 economic crisis. Actually, new trading mechanisms adapted to the technical characteristics of VREs and reflecting their value for the environment and society and that are fair to all energy sectors need to be designed and put into practice.

Section 1.2 provides an overview of the renewable generation physical characteristics, namely, of the temporal variability of these power sources and how the natural complementarity of their primary resources can be exploited to maximize the VRE share into electricity/market environments, without overlooking the security of the energy system. In Section 1.3, different concepts to increase the VRE generation value are presented. In Section 1.4, the challenges of the current market designs for the foreseen nearly 100% renewable power systems are addressed, while in Section 1.5, some of the specifications for new market designs under such levels of renewable penetration are discussed. Finally, in Section 1.6, some final remarks are provided.

1.2 Physical characteristics of renewable generation

1.2.1 Time variability: daily, seasonal, and annual cycles

VREs generation has natural cycles that result from the variability of their primary energy resources, such as wind, solar irradiation, rainfall, and/or melting. Those cycles are explained by the synoptic weather conditions and their interaction with the different topographic characteristics. Since the principle of stable operation of an electrical system requires a constant balance between production and demand, a large contribution from VRE and their reduced predictability (strongly) increases the challenge of managing the power system from its security and robustness point of view. Thus given the power system and markets’ dependency on weather, is crucial to assess the timescale-associated variability with each technology. The renewable energy sources (RES) variability timescales are presented in Fig. 1.1.

Figure 1.1 Renewable energy variability along different timescales.

Solar: Solar irradiance varies in timescales from seconds to years. Part of the seasonal and daily variability depends on the apparent movement of the sun in the sky and it is predictable with high precision with adequate mathematical models. However, for a given site and time, the existence of clouds can induce large variations in the available irradiance with respect to the (theoretical) clear sky irradiation. Although with less impact, aspects such as atmospheric turbidity due to aerosols can also act as a cloud-inducing variability considering irradiation. For photovoltaic systems, additional factors can affect the generation, namely, the temperature of the cell, and obstacles that may shade the system unleashing up/down power ramps [4,5].

Wind: The time variability of the wind velocity is originated by the nonhomogeneity of large-scale air masses. Under certain patterns, local effects like sea/land breezes that stem from the sharp contrast between high temperatures over land and lower temperatures over the sea, atmospheric thermal stratification, local terrain, and surface roughness can have a significant impact on the atmospheric airflow [5]. Therefore, in many geographical areas, there is a marked daily and seasonal wind speed profile. Some sites exhibit pronounced diurnal patterns, which depend on location (Fig. 1.2). In mountain regions, the highest wind speed values are typically observed during the nighttime, while for coastal, flat, and nearshore regions, on average, the daily peak wind speed is observed in the mid-afternoon.

Figure 1.2 Daily wind capacity factor in mountain and nearshore/flat regions. The bars depict the hourly deviation with respect to the daily average value.

Wind power plants convert the kinetic energy of the wind into electrical energy, which they deliver to the grid, and the amount of energy produced is directly proportional to the cube of the wind speed until the nominal conditions are met. Most of the modern wind technology operates with wind speeds normally between 3.0 m/s (cut-in) and 25 m/s (cut-out). During high-speed events, the wind turbine may be turned off for safety reasons. Modern control strategies enable to operate wind turbines above cut-out wind speeds, thus avoiding the curtailment of the available wind power.

Hydro: The capacity of hydroelectric power plants (run-of-river or reservoirs) to produce energy depends, mainly, on the water cycle that affects precipitation and ice melting. Hydropower plants have reduced power fluctuations on hourly and even daily timescales, even when they are not equipped with reservoirs since their primary energy source, that is, the water flow on rivers, does not show perceptive changes within small timescales, reflecting the integrating nature of this renewable energy resource, both in time and space. However, high interannual and seasonal scale variability is common, especially in temperate climate regions. The energy production of a run-of-river hydroelectric (RoR) plant misses the regulating and storage capability of the reservoir, thus it depends directly on the streamflow of the river where it operates which is very dependent on the regional (accumulated) weather conditions. In the European winter, the flow of rivers tends to be higher than average, increasing the production of the RoR power plants; conversely, in the summer, the lower availability of the resource leads to less generation by these power plants. In large power plants equipped with reservoirs, the seasonal variability also dictates the resource availability and, consequently, the amount of energy (in the form of potential energy) that can be stored in reservoirs [6].

The generation variability of the dominant renewable technologies is presented in Fig. 1.3. Wind power is the only one without a clearly diurnal cycle, as the hydrocycle is mainly due to plant control, not to the primary resource. For wind, an autocorrelation above 0.5 is expected for time lags below 1 day. This means that wind power generation has some degree of correlation within a 1-day time horizon, and above that, the correlation value tends smoothly to zero indicating a reduced level of predictability for larger time horizons. Due to its primary resource, solar power exhibits the highest variation in the autocorrelation values, and daily cycles with 24 hours are clearly depicted. The hydrogeneration autocorrelation values show two peak points with an identical period—12 and 24 hours. These peaks translate the energy storage/regulation capability of this technology [6]. The lower autocorrelation values from reservoir power plants, when compared with RoR, expose the controllability capabilities of the hydropower.

Figure 1.3 Autocorrelation of the different RES generation available in the Portuguese power system.

Regarding the common RES technologies under current deployment, geothermal and biomass are the ones that show the lowest variability, due to their independence from weather cycles. However, the potential of these technologies is still limited to very specific locations, thus preventing economies of scale that could lead to lowering the present installation costs. The power variability of those technologies is globally low and they have the capability of contributing to baseload consumption, as well as providing a wide range of ancillary services, contributing to both short- and long-term flexibility needs. Other sources, such as wave/tidal power, have insignificant variability in the second to the hourly timescale, but their fluctuations become relevant on a daily and seasonal timescale [5]. Variability of wave power tends to be quasi-independent from the time of the day, while the tidal power is associated with the periodic variations in gravitational attraction exerted by celestial bodies in water bodies (e.g., oceans, lakes). Although the wave and tidal generation are more predictable than wind and solar power, their reduced expression on power systems has not yet required their variability to be studied in detail.

Climate change and the existence of extreme events (e.g., successive dry years) is expected to have a great impact on the weather parameters previously presented, and, consequently, on the planning, operation, and reliability of power systems as well as in the electricity markets [7]. In the current energy transition, it is paramount to adopt a long-term and holistic approach that considers and anticipates the variability of these generation sources in a cost-effective way, and with acceptable levels of security and reliability [8].

1.2.2 Natural complementarity of renewable energy resources

Understanding the complementarity of the primary resources from VRE power production and their capability to meet the predetermined operational setpoints required by system operators, as well as their capability to participate in electricity markets, is a crucial step toward the decarbonization of electrical power systems without overlooking the security of the energy system and the overall operational costs [8]. This is particularly relevant to enable a smooth energy transition since most of the European power system scenarios to achieve the European Union 2030 and 2050 ambitious targets strongly rely on VRE, such as wind and solar photovoltaic (PV). These technologies are fundamentally different from conventional energy sources due to the stochastic nature of their primary resource. Additionally, cyclic generation patterns exist, for example, the solar power marked daily profile associated to the electricity consumption produces the so-called PV generation duck curve [9], depicted in Fig. 1.4.

Figure 1.4 Net load (consumption minus wind and solar generation) daily profile for Portugal and the impact of different levels of solar PV capacity.

Those renewable weather-induced patterns raise concerns regarding the capacity of existing power systems to absorb the VRE generated power ramps [9,10], that is, the capability of the conventional power plants to accommodate the ramp rate and range needed to fully exploit variable renewable technologies—a concept normally translated by the flexibility of the power system—without (1) increasing the overall costs of the system and (2) reducing the environmental benefits of variable renewables. Associated to the requirements for allocation of extra reserves (i.e., to increase power system flexibility) a strong change in the form in which electricity markets are designed and operate is also expected. Those steep PV and wind ramps can only be compensated by dispatchable power plant technologies or energy storage units, which expose the power systems to a limited number of players in some periods of the day, thus reducing the number of players and the competitiveness of markets. Additionally, increasing the penetration of zero marginal cost variable renewable technologies as solar and wind may lead to a reduction in wholesale electricity prices due to the so-called merit–order effect (see Section 1.4).

Recent studies reveal the existence of VRE generation complementarity on different timescales, from hourly to annual, associated with the diurnal and/or seasonal cycle of solar radiation, which unleashes different weather phenomena across the world, as summarized by Jurasz [11]. For the Iberian Peninsula (IP) regional climate, it was recently concluded [12,13] that the joint operation of wind power with PV would mitigate the wind power variability due to the strong complementarity of the daily profiles of both renewables, promoting the so-called power smoothing effect [14], especially during summer months.

Fig. 1.5 depicts the renewable aggregated generation in Portugal during a 3-year period, including hydropower plants (reservoir and RoR), wind, and solar PV. The daily and monthly correlation values for this RES aggregate (on the country scale) are presented in Table 1.1. It is possible to conclude that, in average terms, on a daily scale, the correlation between hydropower and solar PV is reduced. A similar result is observed for wind power and solar PV. However, on a monthly scale, the complementarity between the solar PV with hydrogeneration is substantial (a cross-correlation factor from –0.42 to –0.48). These results are explained by the hydrogeneration storage capacity, and the planning time of adjustment of hydrogeneration to market conditions. For power systems with very high renewable penetration—typically above 70%—assessing the synergy between hydropower and the remaining renewable technologies (wind and PV) is crucial to maintaining high standards of service quality, namely the long-term security of supply, while minimizing the energy storage requirements, as well as the optimization of its economic performance.

Figure 1.5 Hydro (reservoir and RoR), wind, and solar daily (thin lines) and weekly (thick lines) aggregated generation in Portugal over 3 years.

Table 1.1

1.3 Enhance the VRE value: from the large European to small local electricity markets

The existing competitive and liberalized markets were extremely effective in reducing the costs of electricity by adopting designs adapted to the technical and operational characteristics of the 20th century conventional dispatchable power plants. Those designs rely on the predictability (or predefinition) of the power produced by a dispatchable plant for each time interval, using a (conventional) power plant characteristic normally referred to as its guarantee of power. The challenge for the participation of VRE in markets is they do not offer guarantee of power, that is, one cannot ensure with high reliability that for the next day (for a day-ahead designed market) a wind or PV solar plant will deliver a certain quantity of energy during one hour. Both the time and spatial variability of their renewable resources, as well as the difficulties associated to forecasting them, hamper decisively that objective for VRE plants. That has, as consequence, a strong reduction of VRE market value, since markets design strongly penalizes VRE operation up to a point where it compromises their participation in existing markets. Nevertheless, in recent years novel approaches in the planning, designing, and operating of VRE plants have enabled to reduce VRE variability and, consequently, to minimize the generated power deviation with respect to a forecasted or predefined set-point. This section addresses different approaches that enable to increase the VRE value by taking into consideration their variability and intrinsic stochastic generation features.

1.3.1 Spatial smoothing effects

One of the most basic features for wind and solar PV power production, which can cancel in many cases the rapid fluctuations of VRE power production and minimize the impact of their storage variability, is the lack of spatiotemporal correlation of their primary resources—for example, between wind speed and solar irradiation [5,15]. Fig. 1.6 shows the Pearson correlation and standard deviation of the combined generation (on an hourly basis) considering the existing wind power plants’ locations in Portugal. A high-resolution mesoscale model was used to obtain wind power production time series. For comparison, the exact same locations were used to assess the hourly correlation for solar PV power production based on satellite data.

Figure 1.6 Hourly (A) correlation and (B) standard deviation between wind power plants in Portugal using 2 years of data (2015–16). The same locations were used to calculate the hourly correlation for the solar PV power production.

In the solar PV generation case, the smoothing effect for a timescale of 1 hour is small. This behavior can be explained by longitude/latitude dependency from the primary resource rather than the local effects that are strongly present for wind generation, but not for PV. This physical characteristic of renewable primary sources and the use of concepts as virtual (or hybrid) power plants for several VRE power plants widely dispersed, enables the application of an aggregated dispatch strategy and allows VREs to participate in electricity markets as a single entity (with a certain degree of power control capability). This operational strategy is expected to limit the overall VRE generation forecasting errors, thus increasing their value in electricity markets [16,17]. In recent years, several authors have highlighted the benefits of a statistical smoothing effect to improve the VRE power forecast since their errors in a large geographical area are mostly uncorrelated [17,18]. Consequently, the aggregate wind or solar PV power forecast can provide lower errors when compared to the sum of individual power forecasts, reducing the need to balance at high costs the energy on the reserve markets. In principle, and if grid limitations are not a concern, this would favor the enlargement of electricity markets.

1.3.2 Aggregation and virtual renewable power plants

As a result of the current energy transition, new energy players are emerging—among them, the aggregating agents of variable renewable generation—who contribute to the provision of new electricity services and auxiliary systems [19,20]. These new agents may—in the near future and assisted by information and communication technologies (ICT)—provide renewable and spatially distributed technologies with some control of energy production and/or storage. This technical capability will give VREs some degree of operating flexibility, thus reducing the risk associated with their participation in electricity markets and also contributes to the robustness of the power system as well as its quality of the service.

These days, it is commonly recognized that the aggregate management of several renewable energy sources distributed associated to storage units—in a configuration usually referred to as a virtual (or hybrid) renewable power plant (VRPP) [21]—when integrated into active (smart) networks, will make an essential contribution to an economically optimized operation of electrical systems in power systems with near 100% renewable-based generation.

Although different definitions can be found in the literature, the VRPP can be described as a flexible representation of a portfolio of different renewable energy technologies, capable of aggregating its capacities and creating a single operation profile, offering a certain degree of guarantee of power, and operating as (close as possible to) a conventional power plant [22]. This concept bases its approach on the development of ICTs systems to aggregate, monitor, and control the status of the distributed energy resources, thus increasing security and granting economic sustainability to its integration in power systems and its participation in an energy market (EM) environment [19,21]. Depending on the objective, the VRPP concept can be divided into two main categories [23]: technical and commercial.

The commercial VRPP aims to empower the participation of small- and medium-scale VREs producers in the electricity and ancillary services markets [24], reducing their risk exposure and optimizing their value within a liberalized market environment. In the VRPP concept, the renewable power plants are combined irrespective of their geographical location and the operational parameters. Thus the VRPP participation in the market environments is usually based on the marginal costs and technical characteristics of the combined renewable power plants considering an optimal schedule (and offer) [22,25].

The benefit of the VRPP solution to enhance the economic performance of renewable producers as well as the power system is well-documented in the current literature [20,22]. Fig. 1.7 shows an optimal VRPP’s production profile able to cope with the variability both from local demand and VREs [22]. This particular VRPP is constituted by wind, solar PV, and hydropower plants (both RoR and pumped hydrostorage—PHS). Hydrogeneration, especially PHS units, play a crucial role in regulating VRPP’s production profile. As desired, the combined solution is capable of adapting its production to the local demand profile (almost) independently of the primary renewable resource, operating close to a dispatchable power plant (Fig. 1.7B). Moreover, by shifting the exported generation periods from hours of low spot market prices to higher ones an increase in the VRPP’s total overall value is observed (Fig. 1.8).

Figure 1.7 VRPP (A) production and consumption for 1 week, (B) production profile and spot market prices for 1 week.

Figure 1.8 Marginal value of the solar (PVPP), wind (WPP), and virtual renewable (VRPP) power plants.

Joint participation in the day-ahead market (DAM) allows increasing the value of the energy produced by the VRE power plants by 29% for wind and 2% for the PV comparing to their stand-alone operation (Fig. 1.8). Compared to market average prices, an 8% increase in revenues is obtained by the VRPP in the DAM. This result indicates a significant increase in the total revenue of renewable power plants, even for the PV power plant that, despite producing energy during the day (when DAM prices are higher) can also benefit from the flexibility given by the PHS to shift energy from days of low market prices to higher ones (e.g., from weekends to weekdays).

From the recent Clean Energy for All Europeans Package the RES aggregator emerges as one of the key players in the foreseen energy transition. This new player will act as an intermediary between decentralized/distributed actors and the market by supporting small actors, for example, prosumers, to participate in the electricity market. In general terms, similarly to the VRPP, the aggregator will bring different players within the power system to act as a single entity when participating in the markets or when selling services to power system operators. With aggregation (VRPPs) it will be possible to combine several different units optimizing their operation profiles, costs, and overall value. For consumers and prosumers also participating in VRPPs, the aggregation will have the potential to reduce energy costs, while it reduces the overall power system balancing costs.

VRPPs assume nowadays an extreme relevance, especially when considering the foreseen phase-out of European feed-in tariffs in the near future and the very recent proposals/announcements from the European Commission (EC) within the Clean Energy package, in which the access of the VREs to the grid is proposed as no longer guaranteed.

1.3.3 Emerging of local markets

The reduction in VRE technologies’ costs, especially solar PV, and the advances in ICT and power electronics solutions are transforming electric grids into a decentralized management system due to the active participation of large numbers of electricity consumers—the so-called prosumers. The proliferation of prosumers and new concepts, such as peer-to-peer (P2P) negotiation of electricity, enable decentralized and competitive markets that are not controlled by the conventional electric utility companies, in which their energy surplus could be traded with neighbors, thus allowing maximization of the VREs value. Further benefits are expected with this type of market since (1) the consumers may negotiate the electricity at a fair (local) price; (2) it can promote the VRE deployment in a sustainable form while reducing dependence on subsidies; and (3) they enable the power systems to become more resilient and efficient as the production is located near to the consumption [26,27]. Moreover, local electricity markets can adopt different trading concepts, such as P2P, auctions, or others more adapted to the natural, environmental, and societal characteristics of each region. These markets were regulated at European level in the recent Clean Energy package presented by the EC [27] and are expected to be deployed to their full potential during the next decade