Handbook on the Economics of Renewable Energy 9781800379022, 1800379021

Citation preview

HANDBOOK ON THE ECONOMICS OF RENEWABLE ENERGY

To my wife Cristina and my sons Marcos and Alvaro (Pablo del Río) To my wife Ulrike, my son Adrian and my daughter Gesine (Mario Ragwitz)

Handbook on the Economics of Renewable Energy Edited by

Pablo del Río Senior Researcher, Institute for Public Policies and Goods (IPP), Consejo Superior de Investigaciones Científicas (CSIC), Spain

Mario Ragwitz Director, Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems (IEG) and Full Professor, Brandenburg University of Technology Cottbus-Senftenberg (BTU CS), Germany

Cheltenham, UK · Northampton, MA, USA

© Pablo del Río and Mario Ragwitz 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2023937056 This book is available electronically in the Economics subject collection http://dx​.doi​.org​/10​.4337​/9781800379022

ISBN 978 1 80037 901 5 (cased) ISBN 978 1 80037 902 2 (eBook)

EEP BoX

Contents

viii

List of contributors 1

Introduction to the Handbook on the Economics of Renewable Energy Pablo del Río and Mario Ragwitz

1

PART I  SETTING THE SCENE: THEORETICAL/ METHODOLOGICAL GUIDELINES ON THE ECONOMICS OF RENEWABLE ENERGY 2

Costs and benefits of the energy transition Barbara Breitschopf, Julia Panny and Anne Held

3

Energy system modelling of renewable energy and related energy infrastructure needs Gustav Resch, Franziska Schöniger, Florian Hasengst, Demet Suna, Gerhard Totschnig and Frank Sensfuß

4

Econometric modeling of renewable energy deployment Consolación Quintana-Rojo, Miguel-Ángel Tarancón and Fernando Callejas-Albiñana

11

41

77

PART II  FOCUS ON SOCIOECONOMIC BENEFITS 5

Extended input-output tables to analyze the benefits of renewable energy deployment Santacruz Banacloche, Ana Rosa Gamarra, Natàlia Caldés and Yolanda Lechón

101

6

The socioeconomic benefits of renewable energy projects Ana Rosa Gamarra, Santacruz Banacloche, Natàlia Caldés and Yolanda Lechón

118

7

Green jobs in the Spanish renewable energy sector: an input-output approach Manuel Tomás, Ignacio Cazcarro, Julen Montilla, Cristina Pizarro-Irizar and Iñaki Arto

138

PART III  FOCUS ON COSTS 8

The grid costs of renewable energy deployment Joan Batalla-Bejerano, Daniel Davi-Arderius and Elisa Trujillo-Baute

v

158

vi  Handbook on the economics of renewable energy

9

Guiding the transition: design challenges in decarbonising electricity markets 179 Timo Gerres, José Pablo Chaves, Francisco Martín, Michel Rivier, Álvaro Sánchez and Tomás Gómez

PART IV  FOCUS ON THE BOTTOM-UP: SELF-GENERATION AND COLLECTIVE ENERGY ACTIONS 10

An economic approach to photovoltaic microgeneration Pere Mir-Artigues

11

Enhancing energy democracy and tackling energy poverty by fostering the uptake of renewable energy: the case of Greece Eleni Kanellou, Ifigenia Tsakalogianni, Haris Doukas, and Yannis Maniatis

206

231

PART V  FUTURE-LOOKING PERSPECTIVES: GEOPOLITICAL, RISKS/FINANCIAL AND INNOVATION PERSPECTIVES 12

The economic benefits of renewable energies: a geopolitical perspective Gonzalo Escribano and Lara Lázaro‑Touza

13

Drivers and barriers to renewable electricity technologies: lessons from the technological innovation system approach Pablo del Río and Christoph P. Kiefer

14

15

Analyzing the suitability and role of modern portfolio theory for renewable energy planning Fernando de Llano Paz, Javier Eduardo Afonso Arévalo and Guillermo Iglesias Gómez A mixed-integer linear programming approach for an optimal-economic design of renewable district heating systems: a case study for a German grid Maximilian Sporleder, Michael Rath, Markus Jansen and Robin Mann

251

284

308

340

PART VI  RENEWABLE ENERGY POLICY 16

The economic analysis of renewable energy policies: a general overview and a historical perspective Christoph P. Kiefer, Pablo del Río and Leticia García-Martínez

17

Renewable energy auctions: an overview Vasilios Anatolitis and Jenny Winkler

18

The role of design elements in instrument mixes: the case of auctions and renewable portfolio standards in South Korea Tae-Hyeong Kwon and Pablo del Río

365 392

420

Contents 

19

20 Index

Climate-related development aid for renewable energy projects: an analysis of its trends and role in fostering the low carbon transition in official development aid recipients Cristina Peñasco Conclusion to the Handbook on the Economics of Renewable Energy Pablo del Río and Mario Ragwitz

vii

442 471 475

Contributors

Javier Eduardo Afonso Arévalo, University of A Coruña (UDC), Spain. Vasilios Anatolitis, Fraunhofer Institute for Systems and Innovation Research ISI, Germany. Iñaki Arto, Basque Centre for Climate Change (BC3), Spain. Santacruz Banacloche, Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Spain. Joan Batalla-Bejerano, Foundation for Energy and Environmental Sustainability (Funseam) and Chair in Energy Sustainability, Spain. Barbara Breitschopf, Fraunhofer Institute for Systems and Innovation Research ISI, Germany. Natàlia Caldés, Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Spain. Fernando Callejas-Albiñana, University of Castilla-La Mancha, Spain. Ignacio Cazcarro, Basque Centre for Climate Change (BC3) and Aragonese Agency for Research and Development (ARAID), Spain. José Pablo Chaves, Institute for Research in Technology, ICAI, Comillas Pontifical University, Spain. Daniel Davi-Arderius, University of Barcelona and Chair in Energy Sustainability, Barcelona Institute of Economics, Spain. Fernando de Llano Paz, University of A Coruña (UDC), Spain. Pablo del Río, Consejo Superior de Investigaciones Científicas, Spain. Haris Doukas, National Technical University of Athens, Greece. Gonzalo Escribano, Universidad Nacional de Educación a Distancia-UNED, Spain, and Energy and Climate Programme, The Elcano Royal Institute, Spain. Ana Rosa Gamarra, Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Spain. Leticia García-Martínez, Universidad Carlos III, Spain. Timo Gerres, Institute for Research in Technology, ICAI, Comillas Pontifical University, Spain. Tomás Gómez, Institute for Research in Technology, ICAI, Comillas Pontifical University, Spain. viii

Contributors 

ix

Florian Hasengst, Technische Universität Wien (TU Wien), Austria. Anne Held, Fraunhofer Institute for Systems and Innovation Research ISI, Germany. Guillermo Iglesias Gómez, University of A Coruña (UDC), Spain. Markus Jansen, Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems (IEG), Germany. Eleni Kanellou, National Technical University of Athens, Greece. Christoph P. Kiefer, Fraunhofer Institute for Systems and Innovation Research ISI, Germany. Tae-Hyeong Kwon, Hankuk University of Foreign Studies, South Korea. Lara Lázaro‑Touza, Centro de Enseñanza Superior Cardenal Cisneros, Spain, and Energy and Climate Programme, The Elcano Royal Institute, Spain. Yolanda Lechón, Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas (CIEMAT), Spain. Yannis Maniatis, University of Piraeus, Greece. Robin Mann, Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems (IEG), Germany. Francisco Martín, Institute for Research in Technology, ICAI, Comillas Pontifical University, Spain. Pere Mir-Artigues, University of Lleida, Spain. Julen Montilla, University of the Basque Country (UPV/EHU), Spain. Julia Panny, Fraunhofer Institute for Systems and Innovation Research ISI, Germany. Cristina Peñasco, Department of Politics and International Studies and Bennett Institute, University of Cambridge, United Kingdom. Cristina Pizarro-Irizar, Basque Centre for Climate Change (BC3) and University of the Basque Country (UPV/EHU), Spain. Consolación Quintana-Rojo, University of Castilla-La Mancha, Spain. Mario Ragwitz, Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems (IEG) and Brandenburg University of Technology Cottbus-Senftenberg (BTU CS), Germany. Michael Rath, Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems (IEG) and Bochum University of Applied Sciences, Department of Civil and Environmental Engineering, Germany. Gustav Resch, AIT Austrian Institute of Technology GmbH (AIT), Center for Energy, Austria. Michel Rivier, Institute for Research in Technology, ICAI, Comillas Pontifical University, Spain. Álvaro Sánchez, Institute for Research in Technology, ICAI, Comillas Pontifical University, Spain.

x  Handbook on the economics of renewable energy

Franziska Schöniger, Technische Universität Wien (TU Wien), Austria. Frank Sensfuß, Fraunhofer Institute for Systems and Innovation Research ISI, Germany. Maximilian Sporleder, Fraunhofer Research Institution for Energy Infrastructures and Geothermal Systems (IEG) and Brandenburg University of Technology Cottbus-Senftenberg (BTU CS), Germany. Demet Suna, Austrian Institute of Technology (AIT), Center for Energy, Austria. Miguel-Ángel Tarancón, University of Castilla-La Mancha, Spain. Manuel Tomás, Basque Centre for Climate Change (BC3) and University of the Basque Country (UPV/EHU), Spain. Gerhard Totschnig, Austrian Institute of Technology (AIT), Center for Energy, Austria. Elisa Trujillo-Baute, University of Lleida and Chair in Energy Sustainability, Spain. Ifigenia Tsakalogianni, University of Piraeus, Greece. Jenny Winkler, Fraunhofer Institute for Systems and Innovation Research ISI, Germany.

1. Introduction to the Handbook on the Economics of Renewable Energy* Pablo del Río and Mario Ragwitz

The decarbonisation of energy systems represents a key element of the energy transition which is needed to meet the targets of the Paris Agreement, which was adopted by 196 countries in 2015. Governments agreed to limit global warming to 2 degrees, and preferably 1.5 degrees, above preindustrial levels. Low carbon investments are essential for transforming the energy system into a renewable one aligned with the pledges of the Paris Agreement and more globally with the sustainable development goals. According to the International Energy Agency (IEA 2021), around $5 trillion per year of additional capital investments by 2030 and around $4.5 trillion per year by 2050 are needed only in energy and energy-related infrastructures to achieve net-zero CO2 emissions globally by 2050. Therefore, meeting such a target is contingent upon a large transformation of our global energy production and consumption systems. Renewable energy technologies (RETs), and more specifically technologies which generate electricity from renewable energy sources, are a main pillar of this decarbonised energy transition, together with energy efficiency. “Energy from renewable sources” or “renewable energy” means energy from renewable non-fossil sources, namely wind, solar (solar thermal and solar photovoltaic) and geothermal energy, ambient energy, tide, wave and other ocean energy, hydropower, biomass, landfill gas, sewage treatment plant gas and biogas (EU Parliament and Council of the European Union 2018). According to the International Renewable Energy Agency (IRENA) (2022a), 25% of the reductions of 37 gigatonnes of annual CO2 emissions by 2050, which will be required in order to meet the 1.5 degree target, will be achieved cost efficiently through renewables.1 Currently, the share of renewables in energy production and consumption is modest. In 2020, “modern renewables” only accounted for 12.6% of final energy consumption worldwide (REN21 2022).2 This global average for renewables masks important differences in the penetration levels per region and final use sectors. According to the IEA (2021), world energy supply from renewables amounted to 68.5 exajoules (EJ) in 2020. Three-quarters of this supply came from three regions: Asia Pacific (25.8 EJ), with China accounting for 12.9 of this last figure, North America (12.8 EJ) and Europe (14.3). The contributions from the other regions were: Africa (4.6 EJ), Central and South America (9.5 EJ), Middle East (0.2 EJ) and Eurasia (1.3 EJ). Broadly speaking, renewables can be used in three different sectors: heating and cooling, transport and electricity. They have a different share in the three energy consuming sectors: 11.2% in heating and cooling, 3.7% in transport and 28% in electricity. The focus of this book is on this final energy use sector, i.e., electricity, although the other sectors are also taken into account. In order to meet the Paris Agreement targets, those percentages should increase considerably. According to IRENA, the share of renewables in final energy consumption should 1

2  Handbook on the economics of renewable energy

increase from the current 16% to 79% in 2050 (IRENA 2022a). Focusing on the electricity sector, according to IRENA, the share of renewables in electricity generation should skyrocket from 25% in 2018 to 57% in 2030 and 86% in 2050 (IRENA 2019). A more recent estimation by IRENA shows that the share of renewables in electricity generation should grow even faster from the current 26% to 90% in 2050 (IRENA 2022a). There is a wide consensus that solar photovoltaics (PV) and wind will be the RETs with the greatest diffusion potential worldwide (see, e.g., IEA 2021; IPCC 2022). These two technologies are expected to make the largest contribution to renewable energy penetration in 2030/2050, given their relatively low costs with respect to other electricity generation technologies, whether renewable or not. According to IRENA (2019), annual additions of wind electricity generation capacity should increase three-fold in the next decade (from 109 GW in 2018 to 300 GW in 2030 and 360 GW in 2050) and annual additions of PV capacity should increase four-fold (from 54 GW in 2018 to 200 GW in 2030 and 240 GW in 2050) in order to meet the Paris Agreement targets. Annual solar PV should increase more than three-fold from the current 126 GW/year to 444 GW/year in 2050, whereas annual wind additions should more than double from 115 GW/year today to 248 GW/year in 2050 (IRENA 2022a). Apart from their application in different final energy use sectors, there are other policyrelevant distinctions between different types of renewables in electricity generation: maturity and dispatchability. An analogy between RETs and a human life can be made. Technologies in general and RETs in particular are initially immature: they are imperfect and have high costs. Due to research on those technologies, but also as they diffuse, improvements and cost reductions take place, and they become fully mature. Mature RETs include onshore wind and solar PV. Less mature technologies have a high potential for improvements and cost reductions, whereas this potential is modest for technologies that have already reached a high maturity level. On the other hand, all RETs in the power sector depend on renewable energy resources to produce electricity. In some cases, electricity generation can only take place when the resource is available at the moment when electricity is produced, i.e., when the wind blows or when there are sufficient solar irradiation levels. In case such resources are not available instantaneously, no generation can occur. This is the case with non-dispatchable (also called variable or intermittent) RETs, such as wind and PV. Of course, the electricity generated with those RETs can also be stored in batteries, but large-scale storage is still costly. In contrast, other RETs have an inherent storage feature, i.e., the renewable resource can be stored and used at will. This is the case with biomass (i.e., crops or forest residues) and hydro (water), although the availability of the resource is contingent upon the absence of droughts. It is also the case in concentrated solar power (also called solar thermal electricity) in which a fluid (water, oil and sands) is heated with the sun impacting mirrors and the thermal inertia allows such fluid to be available in order to move a turbine and produce electricity even hours after the sun has set. These dispatchable RETs allow power to be produced “on demand”. Renewable electricity technologies are, thus, a sine qua non in achieving climate change mitigation targets, i.e., to reduce the emissions of greenhouse gases (GHG). However, they provide more benefits beyond decarbonising energy (electricity) sectors. They also reduce the emissions of other local pollutants. Apart from the environmental benefits, renewable electricity provides other socioeconomic advantages such as economic development, employment and investment opportunities. Finally, by having renewable energy projects in their territory, countries can also reduce their fossil fuel dependency and mitigate the risks related to the

Introduction  3

security of energy supply, which is certainly a major policy concern nowadays, particularly in those countries without fossil fuel resources. In addition to those benefits, there are costs. In general, the decarbonised energy transition will not be achieved without significant investment and costs. And also the huge penetration of renewables which will be needed will come at higher generation and system costs, at least temporarily. This is a major concern for policy makers everywhere in the world. Traditionally, RETs have been more costly than their fossil fuel counterparts. However, given the aforementioned dynamic nature of the cost reductions taking place in RETs, some of these have been able to achieve cost-competitiveness with their fossil fuel competitors. According to data from IRENA (2022b), solar PV, onshore wind and hydro are already cost-competitive technologies in terms of direct costs (LCOE) with respect to other renewable and non-renewable electricity generation technologies. However, for some technologies, a high cost gap remains, although the question is whether these technologies will be needed then if such a high cost gap remains. Furthermore, a characteristic of all RETs is their capital intensity. This means that the required investments in these technologies need to be made up-front, and then, when they start to produce electricity, only small variable costs are incurred (operation and maintenance costs, but not fuel costs). The economic implications of this feature are considerable, since those investments need to be financed and this means that those funds need to be available, which might be difficult in some regions of the world and may not be possible in most countries in the absence of public policies which ensure some return on those investments. In fact, the aforementioned required increase in the uptake of RETs would need to accelerate the growth in deployment and penetration of renewable electricity experienced in recent years, which has been substantial, but slower than needed to meet the aforementioned requirements. Renewables generated 28.3% of global electricity in 2021, up from 20.4% in 2011 (REN21 2022). The diffusion of RETs has been driven by demand-pull policies, with administratively set feed-in tariffs and feed-in premiums (ASFITs/FIPs) being the most popular instrument in this regard (see Chapter 16). ASFITs/FIPs have been superseded in recent years by auctions as the dominant scheme for RET deployment worldwide (REN21 2022). However, given their alleged advantages in terms of cost-effectiveness and lower support costs, auctions are the most rapidly expanding form of support for renewable energy project deployment and are becoming the preferred policy tool for supporting the deployment of large-scale projects. The adoption of RET auctions has increased linearly in the last decades, from 6 countries holding auctions in 2005 (IRENA 2017) to 131 countries holding auctions in 2021 (REN21 2022). This compares to 92 countries with ASFITs/FIPs in 2021 (REN21 2022). The above paragraphs shape the background context for this book, which, as the title states, is about “the Economics of Renewable Energy”. The aim of this book is to provide a comprehensive vision on this topic. It applies economic analysis techniques to identify the costs and benefits of renewable electricity development and deployment, inferring policy implications for the future. It does so by adopting a systemic and dynamic perspective on costs: all the costs of renewable energy deployment (i.e., not only the project-level ones, but their impacts at the electricity-system/energy-system level) and factors that affect their evolution over time, with a main focus on deployment policies. In short, the coverage of the book is also broad: it addresses theoretical, methodological and empirical aspects of RET deployment with a clear policy-oriented philosophy. It is also broad in the economics perspectives adopted, since it encompasses different economic approaches and disciplines, with a main basis in environmental economics and

4  Handbook on the economics of renewable energy

innovation economics. Although some economic disciplines are based on different assumptions about human behaviour (i.e., rationality) and, thus, sometimes lead to contradictory results, our vision is integrative, i.e., we believe that different approaches have a role to play, since they capture different aspects of a problem or issue. Such broad coverage and a holistic and comprehensive approach are also reflected in addressing the many sides of the economics of renewable energy: it focuses on both efficiency and distributional issues, on static and dynamic perspectives, on the short-term and the longterm implications, on different levels (macro and micro) and on different renewable electricity deployment scales (including utility-scale, self-generation and community initiatives). Although it is not exhaustive in the treatment of each and every possible topic in renewable electricity deployment (since they are many), we hope that the sufficiently broad coverage of topics and perspectives remains useful for different actors, including researchers, policy makers and practitioners. Written by experts in the field, the book has 19 chapters and is structured into 6 parts. The first part covers the main theoretical and methodological aspects and provides a “framing” for the rest of the book. In Chapter 2, Breitschopf et al. outline a comprehensive approach for assessing costs and benefits for the past and future development of the energy transition in the EU. The chapter briefly outlines the analytical framework to assess the impacts of the energy transition and describes the framework by using the RET deployment in general as well as based on examples. It describes different types of costs and benefits, applies a unified and consistent terminology for costs and benefits of the energy transition and provides guidance for their assessment. It identifies which types of effects arise in line with the energy transition in general, in different economic sectors, regions or technology fields, how they can be assessed on a regular basis, e.g., life cycle or annual, what the system boundaries of impact assessments are, which impacts can be quantified, expressed in terms of indicators and their potential monetarisation, which information and methodological approach is most suitable to get an appropriate picture of costs and benefits of the system that are the focus of the analysis and how to critically assess results in terms of indicators and be aware of limitations. In Chapter 3, Resch et al. assess past developments in energy system modelling of renewable energy and related energy infrastructure prerequisites. Illustrative snapshots are provided, offering a chronological survey of energy system models with a focus on renewable energy and related main thematic fields. Illustrated topics of interest as well as approaches used in corresponding model-based assessments are provided, with a geographical focus on Europe. The chapter provides fruitful insights on how the energy policy debate has emerged as well as how energy system modelling has advanced to provide substance and support decision making in accordance with emerging policy needs. Quintana-Rojo et al. argue that analysing the factors that determine the development and deployment of renewable energy sources (RES) is a valuable academic effort, since it will allow the elaboration of strategies based on the drivers that stimulate investment in installed capacity, as well as the elimination of barriers that hinder the diffusion of renewable technologies. Chapter 4 focuses on one of the analytical approaches that can be applied to analyse this topic, i.e., econometric modelling. Econometric techniques have the benefit that they represent, in an objective and replicable manner, the system of relations that characterises the development of RES. This technique identifies the main variables that have an impact on this development and allows an analysis of the impact on this system of the relations of various technical, economic, political and social scenarios. Thus, the authors provide a critical

Introduction  5

perspective on what econometrics has to offer with respect to other methods, an insightful discussion of some of the problems that may be encountered when using econometric modelling to analyse the drivers of renewable energy deployment and review the state of the art of studies on this topic. Input-output analysis is another technique which has been used in the past to analyse the benefits of renewable energy deployment. In Chapter 5 (Banacloche et al.), the use of inputoutput (IO) to conduct gross or net assessments of the socioeconomic implications of renewable energy deployment is considered. The use of IO methodology extended with environmental and socioeconomic vectors has been proposed as a framework for analysing the sustainability of energy investments, as it is able to capture the total, direct and indirect, impacts of energy investments, considering the global value chains phenomenon both in environmental and socioeconomic terms. This chapter describes the IO methodology and its extensions and discusses the scope of its applications to renewable energy investments. Chapter 6 extends this discussion with the results of several case studies. The second part of the book is, indeed, devoted to the empirical analysis of the benefits of renewable energy deployment. In Chapter 6, Gamarra et al. provide a review and also some concrete examples of case studies of renewable energy investments and policies using multiregional input-output models (MRIO). In addition, a section dedicated to the assessment of external costs of RES technologies is included. The authors show the relevance and wealth of information derived from this type of study with four examples. The objective of the first study was to assess the sustainability impacts of specific renewable energy innovations. The second aimed to evaluate the employment impacts of alternative renewable technologies. The third study conducted a comparative assessment of national programmes of RES investments and the fourth aimed to assess the socioeconomic impact at the regional level in Europe. Besides, the socioeconomic benefits of air pollution mitigation associated with RES penetration in the transport sector were also presented in a final example. One of the conclusions of this chapter is that, when time and resources allow for it, sound and scientific-based quantitative socioeconomic assessments are required to fully understand the magnitude and causality of such impacts in order to inform and guide investment decisions, design policies, implement measures to minimise adverse effects, boost positive impacts and raise awareness. Chapter 7 (Tomás et al.) focuses on one of those benefits from RES deployment, i.e., employment creation. They use a socially extended MRIO to quantify the employment impacts of a decarbonised energy transition in Spain. The authors consider alternative installation cost vectors and analyse the job creation in green investments under different assumptions for two technologies: onshore wind and solar photovoltaic. Their results show that Spain could create between 6.13 and 12.75 jobs for every million euros invested in renewables, depending on the assumptions and scenarios considered. Most of the jobs created in Spain and the rest of the European Union would be direct, while the indirect effects would be stronger in the rest of the world. Finally, we observe that solar energy generates more jobs overall, but wind energy leads to more jobs in Spain (8.89 vs. 8.65, respectively). Apart from benefits, another key element in the economic analysis of renewable energy deployment is costs. Therefore, Part III of this book focuses on this topic. In Chapter 8, Batalla-Bejerano and colleagues start from the assumption that, in order to make decarbonised electricity generation a reality, it is necessary to improve the exploitation of energy resources and the connection of renewable generation systems to the grid. This chapter shows that a successful deployment of power generation coming from variable renewable sources,

6  Handbook on the economics of renewable energy

such as wind and solar photovoltaic, highly depends on the economic cost of system integration. This chapter seeks to look beyond the impact of renewable generation and aims to present the relevance of transmission and distribution networks and the economic fundamentals of power system operation. The chapter highlights the need to redefine the role of electricity networks to be adapted to the future challenges related to energy transition characterised by a large penetration of renewable electricity generation. In the context of the analysis of the costs inflicted by renewable energy deployment and generation on the overall functioning of current electricity markets and their design, Chapter 9 (Gerres et al.) argues that market design revisions are possibly inevitable to meet the three main objectives of electricity systems (trilemma): capacity adequacy, emission avoidance and affordability. Their results from a Spanish case study demonstrate how changing policy scenarios require additional market mechanisms with significant interdependencies among each other. They show how a market mechanism that is not perfectly aligned with the main system objectives, such as renewable payments, is suboptimal to comply with the main objectives. Furthermore, the authors challenge the premise of cheap green electricity for everyone since the remuneration of additional services increases the total system cost. The previous chapters mostly focused on the macro and system level and on utility-scale deployment. Part IV of this book is devoted to the analysis of more micro-level, small-actor initiatives. This includes self-generation and community initiatives. In Chapter 10, MirArtigues addresses the socioeconomic dimension of residential micro-generation, with an empirical focus on Spain. This is a possibility which was created by the enormous reduction of the prices of photovoltaic equipment which started in 2008. The chapter provides a general classification of PV plants, identifies the six factors that influence the design and management of a residential plant (illustrated with a real case) and defines and discusses the three criteria used for the economic evaluation of micro-generation, namely, the rate of return, the avoided cost and the influence of some heuristics (such as social imitation and sufficient satisfaction) that trigger it. In Chapter 11, Kanellou et al. argue that, to successfully mitigate energy poverty (at least in the European Union) while putting citizens at the heart of the energy transition and fostering energy democracy, bottom-up approaches, including renewable energy ones, need to be pursued at a policy level. Collective energy actions can pave the way towards the uptake of renewable energy enabling and incentivising consumers to become prosumers. Due to their nature, collective energy actions can contribute to empowering citizens to participate in the energy market, especially when combined with innovative financing schemes that can secure the necessary funds needed to invest in renewable energy projects. In this chapter, the notion of employing collective energy actions to support citizens to tackle energy poverty leveraging joint energy initiatives is investigated. The principles of collective innovative actions are presented, followed by the notion of community finance and crowdfunding. Then the reasoning for how this approach can be especially beneficial in mitigating energy poverty is given. Finally, the case of Greece is presented. The status of energy poverty in Greece and the legal framework of collective energy actions such as energy communities is given along with the challenges that need to be addressed to foster citizen participation and enhance energy democracy for the case of Greece. Part V broadens the focus and looks into relevant issues for the future of the energy transition in general and for renewable energy in particular, all of them with considerable policy implications: geopolitical aspects, innovation aspects and risk/financing aspects. In Chapter 12,

Introduction  7

Escribano and Lázaro‑Touza observe that the transition towards a renewable regime implies a shift in geopolitical and geo-economic balances. Therefore, this chapter seeks to analyse the international implications of renewables, as well as their linkages with climate geopolitics and the role of the main renewable powerhouses (China, the US and the EU). It also explores the new geographies of renewables and reviews the literature on their geopolitics. The chapter focuses on the bi-directionality linking geopolitics and renewables, which is key to assessing the geo-economic benefits of the latter. Geopolitics has an economic impact because it might make cross-border renewable cooperation economically prohibitive between rivals. However, the renewable regime is thought to be less conflict-prone and more cooperative, offering new opportunities for economic and political cooperation. The chapter concludes that this requires clear and transparent governance mechanisms to avoid replicating the governance and market failures of the fossil regime. In Chapter 13, del Río and Kiefer not only provide an analysis of an important topic (innovation), but they also do so using a relatively new theoretical/methodological framework from innovation studies which has been applied to renewable energy. The authors observe that barriers to the development and diffusion of RETs need to be removed to achieve a high penetration of renewables and the potential drivers of this high share should be activated in order to accelerate their uptake. Therefore, insights on those drivers and barriers are required in order to implement policies which encourage the diffusion of RETs. The analysis of those drivers/barriers can greatly benefit from the literature on technological innovation systems (TIS) which, indeed, has been applied extensively to analyse the drivers/barriers to RETs. Although some reviews on barriers to RETs are available in the literature, none has focused on the insights provided by the TIS approach. The aim of this chapter is to review the literature on the drivers and barriers to RETs using a TIS approach and to identify relevant insights from this literature. However, some weaknesses of the TIS approach when analysing the determinants of RETs have also been identified, which suggest that it should be complemented with other streams of the literature and that some missing aspects should be integrated into the TIS approach in order to provide a more complete framework for the analysis of drivers and barriers to RETs. In Chapter 14, de Llano Paz et al. provide a forward-looking perspective, combining energy planning, risk and financial analysis. The authors propose that the assessment of investment projects linked to energy planning be carried out from the perspective of a long-term investment selection problem. They use modern portfolio theory (MPT) methodology, which is based on maximising the yield-risk dichotomy (or its alternative, minimising the cost-risk) of the asset portfolio for this task. The energy planning problem over the medium to long term for a company, region or territory can thus be conceived as based on an electricity-generating (power plant) energy asset portfolio optimisation model. The intent is to define the efficient management of a portfolio or group of assets with a long service life over time and characterised by the previously indicated uncertainties (technological, economic, regulatory and environmental). In their empirical study of Spain and Germany, the authors show how to apply the methodology to energy planning. Potential trends in investments in specific energy sources can be identified over the upcoming years for a given territory. Their application leads to useful and policy-relevant insights. They show that, if the aim is to achieve an affordable mix, the investments will flow towards onshore wind in Germany and in Spain over the medium term, and towards solar in both countries, over just the medium term for Germany, and over the short, medium and long term in Spain. Risk-effective technologies, such as large-scale

8  Handbook on the economics of renewable energy

hydro, mini-hydro (only in Germany) and biofuels over the short and long term in Spain, are expected to thrive if a risk-averse portfolio is followed. Moreover, regardless of the type of approach chosen, some technologies are expected to be or to remain important, such as coal and gas (over the short run) in both countries and mini-hydro in Spain (over both the short and long run). Some technologies, such as onshore wind in Germany and solar in Spain, have enormous importance in all scenarios, especially in low-cost portfolios, where they reach shares of up to 65%. In Chapter 15, Sporleder et  al. pay attention to the integration of renewables in district heating system (DHS) with a case study for a German grid. The research has shown that large-scale heat pumps combined with waste heat are competitive with heat generation costs of 12 ct (kW h) –1 (including the pricing for waste heat) compared to CHP plants with generation costs of 10.9 ct (kW h) –1 (the gas price of 2021) and 33.7 ct (kW h) –1 (the gas price of 2022). Biomass will also be an economical option in renewable DHS with an estimated price of 15.7 ct (kW h) –1, especially because biomass has the lowest emissions. In 2022, gas-fired CHP plants are no longer competitive due to the rising gas prices. Biogas-fired CHP plants were more expensive than the other two renewable technologies, at 22.1 ct (kW h) –1. However, it might be an option if no heat source at a high temperature level is available and a renewable share of the heat is obligatory. However, shifting to renewable energy systems will require multivalent and sector-coupled systems. The final part of this book (Part VI) is about policy, covering the more general aspects (Chapter 16), discussing key aspects related to the dominant renewable electricity deployment instrument (auctions, Chapter 17), policy mixes in renewable energy deployment, with a main focus on the design of instruments (Chapter 18) and a key international dimension in renewable energy support, with special attention to official development aid (ODA) (Chapter 19). Renewable energy has been supported for decades now, using different instruments and even different policy approaches. In Chapter 16, Kiefer et  al. provide a critical, historical overview of the evolution of public policy support for renewable energy technologies, with an emphasis on the last two decades, deployment support and the solar and wind technologies. An overall framework for the analysis of renewable electricity policy, which includes framework conditions (i.e., targets and policy stability), instruments and their design elements, is provided. The instruments to support renewable electricity deployment are described, and their pros and cons are discussed. The authors also identify the patterns and trends in the adoption of those instruments in the last decades, with a focus on the European Union. Finally, a discussion of the alternatives for the design of different instruments is provided. In Chapter 17, Anatolitis and Winkler provide insights on the application of renewable energy auctions. The authors analyse the history of renewable energy auctions and describe how renewable energy auctions work, their main advantages (e.g., efficient control of renewable energy deployment and lower support expenditures) and their drawbacks (e.g., potentially low realisation rates of projects). Then, the authors present several policy objectives that can be pursued with auctions, among others, supporting cost efficiency and effectiveness, providing an exhaustive overview of auction design elements and their effects on these policy objectives. They discuss the emergence of zero-price bids in offshore wind auctions and approaches to effectively select winners. The chapter concludes with an outlook on the future of renewable auctions and the role of auctions in related sectors, such as the support of green hydrogen. In Chapter 18, Kwon and del Río start from the assumption that the success of instruments for the deployment of renewable electricity depends on how those instruments are designed

Introduction  9

and combined. However, despite the importance of the choice of design elements and the widespread use of instrument mixes, the impact of such choices on the interactions between instruments has not received much attention. This chapter addresses this gap in the literature and illustrates how the choice of design elements for one instrument may affect the outcome of the instrument mix. It also examines the design elements of auctions in South Korea, where auctions, renewable portfolio standards and feed-in tariffs are combined. The case of South Korea shows that the choice of design elements in an auction may have considerable effects on the instrument mix. The results suggest that, although the need to coordinate instruments for improving the success of instrument mixes has often been proposed, this is not only an issue of coordinating instruments, but also of selecting the appropriate design elements of instruments. Finally, in Chapter 19, Peñasco analyses the evolution and characteristics of climaterelated development finance for the generation of electricity with renewable sources since the Copenhagen Accord in 2009, i.e. donors, recipients, technologies and instruments. Then, she revises the available policy instruments in the hands of governments to promote the transition to low carbon economies and the importance of policy mixes. Lastly, using a qualitative comparative analysis (QCA), she explores the simultaneous/complementary role played by those policy instruments and climate-related development finance for renewable energy projects in incentivising the transition to low carbon economies in countries receiving ODA. Her results confirm the importance of different strategies and policy mixes depending on the level of development of the countries analysed. Those strategies imply different combinations of ODA, energy auctions, tax credits, renewable standards and carbon pricing. Understanding these relationships is key to fostering a faster low carbon energy transition with a high share of renewables.

NOTES *

Online appendices for chapters 4, 13 and 19 are available on the companion website at: https:// www.e-elgar.com/textbooks/del-rio. 1. The rest would be accounted for by energy efficiency (25%), whereas electrification (20%), renewable energy-based CO2 removals (BECCS, 14%), hydrogen (10%) and fossil fuel CO2 capture and storage (CCS, 6%) should contribute the rest (IRENA 2022). 2. Within this percentage, renewable heat (biomass, geothermal and solar) accounts for 4.8%, hydropower represents 3.9%, biofuels for transport account for 1% and the share of other renewables (biomass, geothermal, ocean, solar and wind power) is 2.8% (REN21 2022).

REFERENCES EU Parliament and Council of the European Union. (2018). Directive (EU) 2018/2001 of the European Parliament and of the Council of 11 December 2018 on the promotion of the use of energy from renewable sources (recast). IEA. (2021). Net Zero by 2050. IEA, Paris. https://www​.iea​.org​/reports​/net​-zero​-by​-2050. IPCC. (2022). Climate Change 2022. Mitigation of Climate Change. Working Group III Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. IRENA. (2017). Renewable energy auctions. Analysing 2016. IRENA. (2019). Global Energy Transformation: A Roadmap to 2050 (2019 edition). International Renewable Energy Agency, Abu Dhabi. IRENA. (2022a). World energy transitions Outlook 2022. 1.5°C pathway. IRENA. (2022b). Renewable Power Generation Costs in 2021. https://www​.irena​.org​/publications​/ 2022​/Jul​/ Renewable​-Power​-Generation​-Costs​-in​-2021. REN21. (2022). Renewables 2022. Global Status Report.

PART I SETTING THE SCENE: THEORETICAL/ METHODOLOGICAL GUIDELINES ON THE ECONOMICS OF RENEWABLE ENERGY

2. Costs and benefits of the energy transition Barbara Breitschopf, Julia Panny and Anne Held

BACKGROUND The ongoing energy transition and the increasing deployment of renewable energy technologies (RETs) have initiated an intensive debate on the associated costs and benefits. To better understand the dimension of impacts with respect to costs and benefits of the energy transition, a comprehensive and systematic elaboration and depiction of the positive and negative effects of the energy transition, as well as its implementation in relevant sectors, regions and technology fields is necessary. Several studies at national and European level such as Diekmann et al. (2016), Breitschopf and Winkler (2019), Sievers et al. (2019) and Breitschopf et al. (2013) have approached this challenge. This chapter strives to outline a comprehensive approach to assessing costs and benefits for the past and future development of the energy transition in the EU. It describes different types of costs and benefits, applies a unified and consistent terminology for costs and benefits of the energy transition and provides guidance for their assessment. This chapter draws on diverse publications for the analytical framework and impact assessments of renewable energy (RE) deployment (Breitschopf et  al. (2016), Breitschopf and Held (2014), Lutz and Breitschopf (2016), ImpRES1) as well as on technology and sectoral impact assessments. It is important to note that the proposed cost and benefit concept is not a cost and benefit analysis typically applied in project assessments.2 Rather, it presents a framework showing: ●

● ● ●

●

●

What types of effects arise in line with the energy transition in general, in different economic sectors, regions or technology fields. How they can be assessed on a regular basis, e.g. life cycle or annual. What are the system boundaries of impact assessments. Which impacts can be quantified, expressed in terms of indicators and their potential monetarisation. Which information and methodological approach is most suitable to get an appropriate picture of costs and benefits of the system that are the focus of the analysis. How to critically assess results in terms of indicators and be aware of limitations.

This chapter briefly outlines the analytical framework to assess the impacts of the energy transition and describes the framework by using RET deployment in general as examples.

11

12  Handbook on the economics of renewable energy

CONCEPTUAL STRUCTURE OF COSTS AND BENEFITS System Boundaries To properly assess the impacts of the energy transition, we need clearly defined system boundaries. This entails the formulation of the exact research question: which impacts, technologies and sectors are of interest, which perspectives and levels should be considered, and which geographic area, sector or whole economy of a country and time horizon the analysis should cover? The more precise the question, the easier is the definition of system boundaries. The clear formulation of the research questions includes the object of the research: what are we looking at? To define the system boundaries, we identified the following guiding questions to help set the boundaries of the research (system boundaries): ●

●

●

●

●

●

Which impacts do we analyse (research object), that is: what is the focus of the analysis? For example, the focus could be the impact of certain energy policies on society and environment, the electricity system, technological development or the impact of the deployment of clean energy technologies on the economy. Which perspectives and which impact levels do we take into account? This includes the question of whether we assume a rather aggregated perspective considering the overall economy (macro-level), the energy system (meso-level) or the perspective of smaller entities like households or companies (micro-level). What is the scope of the impact analysis: for example, do we take economic, environmental, technological, social and/or societal effects into account? What are interactions with other systems or sectors? For example, is the electricity sector included when analysing the heat or mobility sector, because heat pumps or electric cars link both sectors? Where is the boundary? Which geographic area do we cover with the analysis? For example, do we focus on the local, national, regional or global perspectives? What is the time horizon of the analysis? For example, do we include effects from past and present deployment or future developments and deployments?

Defining the Perspectives or Levels of the Analysis When looking at potential impacts of the energy transition, we distinguish between three main levels or perspectives, at which different actors are affected (ImpRES3, Diacore4, Perception5). However, to identify and potentially quantify the effects, we need a reference to which we compare the “energy transition situation”. Such references are often counterfactuals, baselines or alternatives. A comparison between a reference and an “energy transition situation” displays the so-called “net” or “additional” effects arising due to the energy transition. They occur at three levels: ●

Actor-specific effects take a micro-economic perspective. They encompass costs or benefits for individual agents or groups. These costs and benefits accrue from the energy transition and stand in contrast to the effects of a reference situation, i.e. an alternative

Costs and benefits of the energy transition  13

●

●

scenario to which the case in focus shall be compared. In addition to the question of the overall costs, the actor-specific perspective analyses who has to pay for the costs and who may obtain benefits. The question of the distributional effects is therefore very relevant for the actor-specific perspective. The distribution of these effects is the result of policies, regulations or institutions that direct how the system-related additional costs or benefits are distributed among consumers, producers, suppliers, aggregators or other service providers, or in economic terms among households, firms and the state. Some agents belong to more than one actor group, thus summing up the effects across actor groups is not possible. For example, private households may pay a levy for RET deployment. At the same time they could also operate a roof-top PV plant and may receive a policy support-based remuneration above generation costs that results in additional profits. System-related effects encompass all direct and indirect costs and benefits of the energy transition within a defined system compared to the reference system. Direct costs include all the costs that are directly related to the energy system such as the installation, operation and maintenance of RE technologies. Indirect costs are caused by integrating renewable energy into the existing generation system, for example grid extension costs, balancing costs, etc. Benefits from RET use arise, for example, as a result of avoided greenhouse gas (GHG) emissions and air pollutants. Furthermore, system-related effects reflect the costs of input factors based on market prices of labour, capital, natural resources and materials. Finally, these costs are identified from a system perspective without taking into account any policy support payment. Distributional effects at the system level may include an analysis of burdens for energy carriers of the different sectors and applications including levies or other price components used to refinance additional costs arising from the energy transition. Macro-economic effects are measured at the macro-economic level and should strictly reflect net effects in an economy. This means that we compare economic key data such as employment or GDP of a counterfactual situation to the key figures of an “energy transition” situation. Thus, net effects (net employment, GDP), for example clean energy technologies, encompass all positive and negative effects of all sectors in an economy. To do this in a comprehensive way, we need macro-economic modelling that includes system-related costs and benefits as well as actor-specific effects, of for example RET deployment, and compares them with a reference situation. However, in reality so-called gross effects are commonly applied. They refer to a specific sector, for example the RE sector, and only comprise the effects in industries that are directly related to RET activities such as manufacturing, operating, construction, research, etc. Further, they provide information on investments, avoided imports, turn-over, gross employment, etc., and ignore potential negative effects “outside” the RE-related industry. They do not rely on a scenario comparison (with and without RET deployment). Therefore, these are (positive) sectoral effects, which do not reveal whether the overall effect of RET deployment on the whole economy is positive or negative. Distributional effects are less important for the macro-economic perspective, as these are levelled out.

It is worth noting that environmental or social effects could entail so-called second-order effects in the economy. For example, environmental effects such as air pollutants have implications for health, which in turn could affect production or income. To get a comprehensive

14  Handbook on the economics of renewable energy research question: effects of the energy transition

level

micro

macro

system, sector

actors

economy, society

impacts

effects

meso

perspective

- compared to a reference situation -

system-related effects: monetary

costs and benefits for systems, sectors

actor-specific effects: monetary and non-monetary costs and benefits of actors

positive impact: returns, increase efficiency, reduced costs or expenditures, etc.

macro-effects: socio-economic, societal effects

negative impact: costs, burdens, usage of resources, reduced benefits, income, etc.

Source:   Breitschopf and Diekmann (2011), adapted

Figure 2.1   Analytical framework picture of the impacts, we should identify and integrate all effects at the different levels (system, micro-economic and macro-economic) in a comprehensive macro-socio-economic assessment model. The analytical framework for identifying the boundaries of the research question is depicted in​Figure 2.1. It highlights which impact levels we can include in the assessment, which perspectives we may take and what the macro-effects include. Furthermore, it stresses that the effects at the meso- and micro-level trigger the impacts at the macro-level. Reference Situation When analysing the impacts of projects, policy instruments or transition within sectors or society, we need a counterfactual situation to display the actual impacts of the research object. The reference situation should display how the situation would have been without the research subject. For example, evaluation studies of energy efficiency measures6 use different approaches such as before/after comparison, control groups, trends in energy efficiency, minimum standards or modelling results to determine the reference situation. Reference scenarios are based on assumed developments of the energy system without RET or support policies. They are developed with the help of energy models. When the research objects are RETs, conventional or alternative technologies meeting similar needs as RETs may be used as the reference. Alternatively, a reference situation might also depict the effects of a different RET, e.g. solar PV instead of onshore wind. Overall, reference situations encompass:

Costs and benefits of the energy transition  15 additional contracted reserves provided by generators and consumers

additional generation costs

additional system costs

additional balancing and profile costs additional costs for grid infrastructure transaction costs

additional reserve costs (forecast errors)

balancing services (generation based) balancing services (capacity based)

additional transaction costs for balancing implementation costs

additional frequency response costs

additional demand response costs additional capacity provision (local) additional costs for planning, procuring, monitoring, data exchange, contracting, …

• mandatory frequency response of generators to changes in frequency (primary) • contracted frequency responses for low or high frequency events • contracted frequency control by demand (demand interruptions up to 30 min)

requirements of policy: policy monitoring, evaluation, prognosis

Figure 2.2  Overview of additional costs ●

●

●

●

Before and after comparisons: the “after” situation includes the same environment of the research object as the “before” situation. But due to trends in technology, changes in behaviours, society and in the economy, the environment of the two situations might differ not only because of the research object. Thus, the difference in the situation might not fully reflect the actual additional impacts of the research object. Control groups: in such a case, subjects such as households, individuals, firms and communities with similar characteristics are selected and compared to those that are targeted in the analysis and affected by the research object. Since these two groups are not completely identical, the difference between them does not reflect the actual additional impact of the research object. Reference technology: technologies that are supposed to be replaced by or fulfil the same purpose as the research object could be used as reference. For example, the replacement of pumps by those with a higher energy efficiency or of conventional heating and electricity generation technologies by RETs. In such a case, the research object is the technology. Trend: trends are applicable when policy instruments or projects aim at the acceleration of a trend. For example, buildings become more energy efficient over time even without policy support due to comfort reasons, the gradual diffusion of efficiency-independent standards in buildings, the cost of energy and retrofitting, etc. When analysing the impact of energy efficiency measures, this trend in the energy efficiency of buildings could serve as a reference. Similarly, this applies to RETs. The trend of RET deployment could serve

16  Handbook on the economics of renewable energy

●

●

as a reference to measure the impact of RET support policies that aim at increasing the deployment of RETs. Minimum standards, current shares or target achievement: when analysing the impact of support policies, existing minimum standards are the baseline to assess the additional effects of the “new” project. For example, policies supporting the retrofitting of a building should have impacts beyond the minimum standard. Otherwise they fail to be additional. This applies also to RETs, where the current share of RETs in electricity generation serves as a baseline, to which “new” support measures or RET projects are compared. Modelling: to get a reference situation based on modelling is a complex and time-intensive approach if models are not already available. Often models are applied to assess the energy consumption of buildings based on key characteristics such as size and materials, or of industries based on resource efficiency and other technology assumptions. Analogously, energy system models are applied to create a reference system that is based on reference technologies, restrictions and policies.

In practice, combinations of reference situations have proved to provide rather reliable results. Especially when applying models, some assumptions or restrictions in models rely on trends, reference technologies or minimum standards. Some evaluation studies assess the additionality of the research object based on a combination of before-after with control groups.

COST AND BENEFIT CATEGORIES: THE CASE OF RENEWABLES In this section, we apply the analytical framework to identify and assess the impact of the current energy transition using RET deployment as an example. In a first step, we identify subject and research question of our analysis, including the scope of the energy transition – geographic coverage, time horizon, technologies and sectors coupled to the electricity or energy system. Second, we differentiate between the three levels of impact and outline the additional effect on the components of the electricity system, actors and the whole society. Third, we select an appropriate reference situation – system or technology – to assess the additional effects. The term additionality is important for all three types of effects as it emphasises the actual costs and benefits that accrue only because of the energy transition. The research subject is renewable energy deployment, for example in Germany, in the current year and with a potential outlook of ten years into the future. This includes all renewable generation technologies in the electricity and heat sector. For the reference, we apply different approaches: for the system level, we use reference technologies and modelling of an energy system based on cost optimisation, and without regulation and support policies for renewables. The reference system modelling provides inputs to assess the resulting effects on the micro-level as well as for the macro-economic modelling of a reference situation. We outline the effects and applied references at the three levels in the following paragraphs. Actor-Specific Effects of RET Deployment Actor-specific effects are distributional effects from the actor perspective such as firms, households, selected actors or the public sector. They depict changes in costs, prices, revenues,

Costs and benefits of the energy transition  17

quantity or quality of services and products as a result of the deployment of RETs. These actor-specific effects may represent beneficial effects for certain actors in the system, while having negative effects for other actor groups. One and the same actor might face benefits as well as costs as a result of different impacts. These changes in costs and benefits for actors might be driven by policies such as the Renewable Energy Act in Germany (remuneration scheme) or the Fond Chaleur in France (RET investment support) that determine how the system-related costs and benefits are distributed among actors, or by regulations regarding infrastructure and markets that affect structures of systems and hence the distribution of costs and benefits, or external factors. From the specific actor perspective, they do not necessarily reflect the use of resources for energy provision. Based on a classification of distributional effects in the economic literature (see Fullerton 2009), the actor-specific effects of RE policies can be classified into six types (see Breitschopf and Diekmann 2013): ●

●

●

●

●

Effects on consumer surplus: changes in power prices through surcharges or price increases due to renewable support payments for power lead to an increase in retail electricity prices and, hence, reduce consumer surplus. Moreover, changes in the generation structure such as an increased use of renewable energy technologies – except biomass technologies – which have close to zero marginal costs, lead to a shift in the supply curve, resulting in lower electricity prices in the wholesale market, and, hence, increase the consumer surplus. Effects on producer surplus: tariffs or premiums paid for RE electricity augment the total producer surplus of investors in renewable energy sources (RES). At the individual level, some producers benefit more than others as their production and cost conditions differ depending, for example, on the location. Consequently, individually differing margins may occur. Scarcity rents: instead of providing direct RES support to reduce CO2 emissions in the power sector, a CO2 price implemented either through a tax or an emission trading scheme (ETS) can be established. The prices for the permits traded in the ETS reflect their scarcity, which is influenced by the quantity of RE-based energy use. Capitalisation effects: these display the indirect or consequential effects of RET deployment for certain actors and encompass changes in the market value of assets. Examples are changes in the value of land that is eligible for wind power generation, or surfaces (roof tops or land) for solar power. Further examples are the impact on the value of real estate located close to a nuclear or coal firing plant that will be given up. In contrast, the value of real estate is likely to decrease if a wind farm is built. These RET-induced changes in asset values cause positive or negative rents for a limited circle of actors; this means asset owners. Changes in utility: besides price effects on assets, RET deployment affects the individual wellbeing of actors, or in economic terms their utility. Examples are an increase or decrease of emissions of GHG, air pollutants, noise of wind power plants, changes in the landscape, etc., that entail individually differing marginal utilities. The utilities differ across groups ranked by their wealth, gender, age, region, urban-rural households, ethnicity, etc. Since utility depends on individual preferences, a common measure for these changes in utility is not applicable. Furthermore, the number of affected people varies according to the type of emissions. For example, emissions of GHG have a global

18  Handbook on the economics of renewable energy

●

impact on climate change, while air pollutants are more local. Although assessment approaches for damage costs of CO2 exist, the actual non-monetary effect on utility for each individual significantly deviates from the modelled damage costs. These individually perceived utility changes are distributional effects that will not be further discussed. Transitional effects and others: in the phase of transition, distributional effects are of a temporary nature at all levels. This means they will fade, change or be compensated over time. Examples are impacts on technological learning, competitiveness of firms, prices of inputs, demand for inputs, infrastructure, etc. In the context of RETs, the focus of this study is on public R&D spending where recipients could financially benefit from public support through the generation of knowledge and new products.

Given this characterisation of distributional effects, we differentiate the actor-specific effects into direct financial effects and indirect price effects and non-monetary effects. ●

●

●

Direct effects such as direct payments to generators or from consumers or governments are called financial effects. For example, some economic actors such as private households are charged a levy or surcharge to pay for the additional costs of RET deployment, while producers or investors of RET-based generation facilities benefit from certain support policies such as remuneration schemes or investment grants. Indirect price effects arise through changes in the energy price or technology price as a result of the market mechanism; this means changes in the quality of goods, their supply and demand. Examples are capitalisation and scarcity rents as well as a higher consumer surplus through declining electricity prices due to the increasing share of RETs in generation. Further, transitional effects such as subsequent impacts of RET deployment on technology costs through learning effects and economies of scale are counted as indirect effects as well. Non-monetary effects encompass impacts on the individual utilities that arise due to nonmonetary effects, mainly due to changes in the environment. They cannot be directly expressed as utility is an abstract construct. Instead proxies are applied such as the willingness to accept or pay for.

Table 2.1 outlines the relevant distributional effects of RET deployment. RE policy support costs The deployment of RETs has been supported by a variety of policy instruments ranging from price- or cost-based support to quantity-based support. If the use of RETs causes additional costs at the system level, these costs must be borne by someone. How these costs are financed is determined by regulations. The financing of these RET promotion schemes relies on two basic financing schemes that can be combined as well: (i) Consumer-based financing (ii) Budget-based financing

Costs and benefits of the energy transition  19

Table 2.1  Overview of distributional effects Type of effect

Actors

Financial

Policy support costs

Consumer, • Consumer-based burden sharing producer, state • Public budget-based burden sharing; this means resulting policy support costs are financed through the state budget • Special exemptions/equalisation schemes for selected consumers such as energy-intensive industries • Investment grants, subsidised interest rates, tax credits

R&D support for technology

Technology developer or provider

Price

Merit-ordereffect

Description

Direct monetary transfer to compensate for costs that cannot be covered because of non-realisable rents (in the short term) due to spill-overs and non-exclusion of uses Change of market prices due to changes in the merit order of the power supply (changes in the order of the technologies in the generation portfolio)

Consumer-based financing Consumer-based financing refers to the financing of RET deployment by final consumers without any support from public budgets. As the power and heat sectors have different features, the applied policy instruments differ. In the power sector, the feed-in premiums in Germany are a good example of an RE policy instrument financed by the final consumers (households and firms): the difference between feed-in premiums paid for electricity from renewable energy sources (RES-E) generators and the market wholesale prices at the respective time plus all additional transaction and balancing costs sum up to the policy support costs. These policy support costs are divided by the total amount of power consumed (less privileged consumption). The resulting surcharge per unit of consumed electricity (€/kWh) is added on top of the electricity price of final consumers. Part of the industry may enjoy privileged consumption, and thus pay a lower surcharge, while non-privileged consumers pay more. In a quota system in which RE certificates are traded, the additional costs of RET deployment for actors are reflected in the certificate prices. In a functioning market the costs for certificates equal the additional costs of power generation with RES and the certificate costs are priced at the market price for electricity. Subsequently, final consumers pay for the RET use. To assess the additional costs for consumers, the traded certificate prices could be multiplied by the respective trade volume. This reflects the sum of surcharges paid by consumers. Dividing the sum by electricity consumption displays the average surcharge per unit electricity. To what extent these certificate costs are actually handed through to the different consumer groups – industry and households – cannot be assessed in the framework of this study. For an in-depth description and analysis of support schemes for the deployment of RETs, see Chapter 16.

20  Handbook on the economics of renewable energy

In the heat sector, obligations like emissions standards or quota w/o certificates are common instruments to induce demand for renewable heating technologies (RET-H). As heat generators and consumers are in many cases identical, the additional micro-economic costs are the same as the system-based additional generation costs, if no further support instruments are applied. If RE certificates are traded, the certificate price multiplied by the heat generation plus any other costs that accrue due to trade gives the costs for final consumers. Budget-based financing Budget-based financing refers to the financing of RET deployment by the government or state budget and can be used to finance feed-in premiums, investment grants or tax credits for RE generation or capacities. Private households and firms will be indirectly affected as public spending for other activities decreases or taxes increase to compensate for RET-related expenditures. In contrast to feed-in schemes, which can be financed either by consumers or the public budgets, the financing of a quota obligation is typically consumer-based and directly affects consumers. In the power sector, guaranteed prices or premiums may be completely or partly financed by the public household. In such a case, the same methodological approach as in the case of consumer-based burden sharing should be applied: the additional micro-economic costs should be disclosed in the respective public accounts. Other policy support instruments comprise financial support for investments or support during operation. This can be achieved either by transfer payments from the public household to investors or operators to cover part of the additional generation costs or by tax credits for generation or investments.7 The monetary volume of both support instruments should be disclosed in the public accounts and annual subsidy reports. The total support of subsidies and tax credits provides an estimation for the additional micro-economic costs paid by the public household. In the heat sector, publicly financed investments or generation via subsidies or tax credits are common measures to support RET use. Thus, the recipients of such support instruments receive a financial advantage paid by the state that reduces their overall financial burden due to RET use. To assess these additional public micro-economic costs, the same procedure as for the power sector can be applied. Special equalisation scheme for industry Consumer-based burden sharing applies not only for private households but also for any other consumers such as industries and services. The consumer-based burden sharing reflects an interference of the government in the power market. This may strongly impact the energyintensive industries’ competitiveness, as the burden for RET deployment in other countries might be lower. In order to reduce the market interference and avoid market distortions, governments have a special surcharge or schemes for industry: they offer a reduced renewable energy surcharge to energy-intensive industries that are exposed to strong international competition. But, in turn, this leads to domestic market distortions, as not all industries can benefit from this special equalisation scheme. This results in an increased burden for the remaining actors. To assess the effect of this special scheme, the total exemptions that are shifted to and finally paid by non-privileged consumers are estimated on a yearly basis. Merit order effect and market value The generation of electricity from RE sources affects the market prices of power as the variable generation costs of most renewable energy power plants (all except biomass power plants)

Costs and benefits of the energy transition  21

are close to zero. Hence, in an energy-only market, where the marginal cost of the last operating generation plant sets the market price, the supply curve shifts to the right. The larger this shift is, the more low-variable-cost RETs enter the market. Thus, the market entry of RE generation technologies tends to lower market prices. This price-decreasing effect is called merit order effect, as the order of operating power plants changes with increasing RET share. As this effect depends on the current load profile and available supply, it can only be assessed with an energy sector model. The electricity market price of a system with RE and without (a few) RETs should be modelled and compared. The difference between the price or traded volume with and without RETs discloses the merit order effect, either as total (€) or specific effect in €/kWh. In order to undertake the modelling approach, detailed data on costs, capacities, availability, etc., of power plants as well as the load profile and constraints must be taken into account. Moreover, the design of the reference scenario should reflect realistic assumptions on the dispatched generation plants. Besides the impact on electricity wholesale prices, the “market value” of variable RES may deviate from those of dispatchable technologies. Depending on the time of feed-in, market prices may be higher or lower than on average. Thus, some RET generates power in peak times when market prices are generally higher, whilst other RET feed-in into the grid in times of lower demand. In particular, the increased value of RET power should be captured in the calculation of policy costs which take into account the difference between the received market price and the paid tariffs. Similarly, in a quota system with functioning certificate markets, the policy costs are the expenditures for certificates. Policy costs or quota systems may be corrected by the difference between the market value of RETs and the average electricity price. Research and development (R&D) support expenditures for RETs Government spending for research, development or demonstration of new technologies is one major support for technology development, knowledge generation, networking, exchange of know-how, etc. It is classified as a technology push instrument in innovation economics. As there are spill-over effects of R&D and no exclusion of competitors from using research results, public support is needed to incentivise research and development. However, R&D support could also be considered as a distributional effect as it lowers the research costs of technology providers and opens windows for temporary rents or profits due to advances in competitiveness. Data on R&D support are for example published by national statistical offices or ministries (see ISI et al. 2010, 2013). System-related Effects of RET Deployment System-related effects occur for all elements of the system, comprising institutions, network infrastructure, generators, suppliers and service providers of the energy system. The term “system” may either refer to the energy sector as a whole or, on the basis of a final energy sector, to, for example, the electricity system or it may be broken down to the technology level. While some costs can be separately depicted at the technology level, for example generation costs for selected generation technologies such as solar power, there are cost categories that are more difficult to assign to a single RET, such as grid infrastructure costs. In this case, the additional costs of a RET electricity system as a whole should be considered.

22  Handbook on the economics of renewable energy

The system-related effects comprise the costs and benefits that arise from integrating RETs into the energy system and are assessed without taking into account taxes or subsidies or other policy-induced transfer payments for investment or generation. The main principle for determining system-related costs and benefits is quite straight-forward: it compares the costs and benefits of two systems: ●

●

A system with a significant share of RETs (called RET-scenario), that is typically achieved as a result of policy efforts (regulations and support policies) to trigger the use of RETs. A reference system, which is typically based on a cost optimisation approach and includes nuclear and fossil fuels, and probably a lower share of RETs. This reference system is often called reference scenario, business-as-usual-case or baseline.

When analysing the additional costs and benefits, it has to be ensured that no double counting takes place. If CO2 costs (in terms of CO2 prices) are internalised in the electricity generation costs, then the benefit from avoided CO2 emissions must be reduced by the internalised CO2 costs (see also Breitschopf and Diekmann 2010). The same principles for avoiding double counting apply to costs related to balancing services. If electricity generation costs include investment, fuel, operation and maintenance expenditures required to balance and back up the power system, expenditures for balancing cannot be separately accounted for. Figure 2.2 provides an overview of system costs, with a special focus on balancing and profile costs (see next section for a precise definition of generation, balancing, profile, grid and transaction costs). Decreasing technology costs partly occur due to learning effects. As technology costs are fully reflected in investment expenditures, they are part of the levelised cost of electricity (LCOE) and reduce the additional generation costs of RETs. Therefore, no separate accounting of the effect is required if the focus of the analysis is the current year. If a dynamic perspective is taken, technology costs depending on cumulative deployment (diffusion) should be integrated into the calculations of future generation costs. Similarly, increasing exports or lead market shares are additional effects that are captured in the turn-over of manufactures. Provided that they affect GDP or employment, they are considered as macro-economic and not as system-related effects. Costs of RET deployment The types of additional system costs, occurring at the system level, depend on energy system design and could therefore differ from system to system. Additional system costs can roughly be differentiated into two types of costs, additional direct and indirect system-related costs. The additional direct costs can be assessed at technology level and reflect only the costs arising from the generation of heat or power with RETs reduced by the avoided costs of conventional generation. Thus, direct system-related costs include the cost difference between the replaced conventional and the new renewables-based technologies. A positive difference indicates higher costs of fossil generation, and thus positive effects of RET deployment, while a negative difference shows additional costs of RET deployment. Provided that most RETs are still more cost-intensive than most conventional generation technologies, we cover systemrelated costs or benefits in the cost category. Additional indirect costs may also include costs that are not directly linked to electricity generation and in some cases cannot be directly allocated to a certain technology. As far as the electricity sector is concerned, these indirect costs arise as they refer to system

Costs and benefits of the energy transition  23

components that are required to integrate RETs into the system and maintain a stable power system. They include for example additional costs for grid infrastructure adjustments (see also Chapter 8). As heat generation is more decentralised, mainly based on non-intermittent RES or combined with non-intermittent energy sources (such as gas and solar thermal) and often located close to demand sites (choice of sites is less dependent on resource availability), indirect additional costs such as additional balancing efforts can be neglected. The different types of additional system-related costs for power generation are depicted in Table 2.2. Additional generation costs Generation costs are calculated on the basis of LCOE. The cost components of LCOE include expenditures for investment, fuel, operation and maintenance of the generation plants that are installed to provide energy and not to ensure the stability of the system. Such generation costs are assessed for the RET scenario as well as for the reference scenario. Power sector To assess the generation costs at the system level, the LCOE are calculated for each generation plant and weighted according to their respectively supplied quantity of power or heat. The difference between the generation costs of the two systems (RET system – reference system) shows the additional electricity generation costs at the system level. adGCsys = å RET ( LCOE RET * QRET ) + åsup ( LCOEsup *Qsup )



- åref ( LCOEref * Qref )

(2.1)

adGCsys: additional generation costs at the system level LCOE: levelised costs of energy in €/kWh Q: quantity of energy in kWh sup: conventional generation (plants) that is necessary to provide the required generation capacity in an RE-based electricity system ref: reference system, which is typically based on a higher share of fossil fuel or nuclear-based generation technology RET: renewable energy generation technology At the technology level, the LCOE from conventional sources of a reference system are weighted according to their respective share in the energy mix that has been replaced by the respective RET. The difference between the LCOE of RETs and their supporting conventional generation technology and the weighted fossil energy mix multiplied by the quantity of RE and of supporting generation gives the additional generation costs per RET. The calculation is as follows:

adGCtech =

(( LCOE

RET

)

+ LCOEsup ) - åref ( LCOEref * weighted ) * ( QRET + Qsup ) (2.2)

adGCtech: additional generation costs at the technology level

24  Handbook on the economics of renewable energy

Table 2.2  Additional system costs in power generation Additional effects

Sector

Description

Generation costs (direct costs)

Heat, power

Costs arising from electricity and heat generation:

Balancing costs (indirect costs)

Power

Balancing costs occur due to the variable generation of electricity from solar and wind power. In case of deviations from the scheduled generation, balancing may either increase or decrease the electricity fed into the grid. Additional balancing costs are the difference between the balancing costs of the two electricity system scenarios (reference and with RETs)

Profile costs (indirect costs)

Power

Refer to the back-up capacity to ensure system and supply security. According to Ueckerdt et al. (2013) profile costs occur due to the following effects:

(i) Difference in costs between RE-based and conventional energy technology-based generation (reference), where the reference could comprise a mix of conventional technologies that is replaced by a RET, such as wind power replacing a mix of fossil technologies (lignite, natural gas, oil, coal) (ii) Difference in costs between a mix of RE and conventional generation technologies and a mix of conventional energy generation technologies (reference)

• A potential increase of average generation costs of the residual load as a result of RES-induced decrease of utilisation of conventional power • Additional capacity of dispatchable technologies required due to the lower capacity credit of non-dispatchable RES such as wind or solar to cover electricity demand at peak times and simultaneous low RES generation • Potential curtailment of electricity in times of overproduction from RETs represents a foregone revenue (benefit) of generators or in case of compensation payments costs for the system Grid costs (indirect costs)

Power, heat

Reinforcement or extension of transmission or distribution grids as well as congestion management including redispatch required to manage situations of high grid loads in the electricity sector. Additional costs are assessed via a comparison of infrastructure costs in the reference and RET scenario (modelling approach) Investment and operation expenditures for additional heat and hydrogen grids that are needed under the RET scenario (compared to the reference scenario) (modelling or before/after approach)

Transaction costs (indirect costs)

Power, heat

Market transaction costs: additional forecasting, planning, monitoring, procuring power, establishing trade, contracting, data exchange, etc., due to the use of RETs Policy implementation costs: administrative costs to implement RE policies or fulfil data provision requirements (for example accounting, approvals) Both types of transaction costs are compared to a before-after situation of RET deployment

Source:   Modified from Breitschopf and Held (2014)

Costs and benefits of the energy transition  25

The weighting of generation technologies from the reference system reflects an estimated substitution rate, that is, the extent to which the RET replaces the respective technologies from the reference system. Adding up the additional generation costs per technology shows the additional generation costs at the system level. Heating sector The quantification of additional costs arising from using RES in the heating sector is more challenging than in the power sector, as the heating sector is characterised by a large variety of heating systems and a predominantly decentralised heat generation. The starting point of the assessment at the technology level is the quantity of heat from renewable energy sources (RES-H) generated per renewable energy source or technology and used per building type. The difference between the LCOE for RETs and conventional heating technologies per building type shows the additional generation costs per technology. The calculation is depicted in Equation 2.3. As one RET for heat replaces different shares of conventional heating technologies (reference technology), a respective reference heating technology mix8 should be applied. The differentiation into building types allows for a better assessment of the total heating costs.

adGC REsys tech g = ( LCOE REsys g - åref * ( LCOEref g * sref g ) * QREsys g (2.3)

adGCtech: additional generation costs at the technology level LCOE: levelised costs of energy (heat and power) in €/kWh Q: quantity of energy (heat and power) in kWh by each RET and g s: share of fossil-based energy that is replaced by the respective RET ref: reference heating technology, such as fossil fuel-based generation technology g: building type (single family home, flats, service and producing sector buildings) REsys: renewable energy generation technology plus its accompanying conventional heating technology, if there is a combination of technologies To assess the additional annual generation costs at the heating system level, all expenditures for investments (annuity of investment expenditures), fuel, maintenance and operation for all technologies of an RET-based and a non-RET-based system are added and compared similarly to the calculation shown for power technologies. Additional balancing costs Most RETs used for the generation of heat are not variable in their output and available throughout the year, or they are combined with a generation system that provides heat on a reliable basis at any time. For example, when using solar thermal energy the technology is always linked to non-intermittent energy sources such as gas, oil or biomass. In central systems, intermittent generation technologies are combined with non-intermittent technologies or storage systems, and both are part of the total investment of a district heating system and thus part of the cost calculation. Consequently, in decentralised or centralised heating systems, costs to ensure permanent and reliable heat generation are already included in the generation costs and, thus, need not be accounted for separately.

26  Handbook on the economics of renewable energy

However, in the power sector, some RETs are variable and stochastic in their electricity output. Due to the related difficulties in reliably predicting the electricity output they are not fully dispatchable. Forecasts on RE-based (mainly wind and solar) power generation are subject to errors, which have to be “balanced” in the short term to guarantee that total electricity supply matches the demand profile. To compensate these errors and maintain system stability, for example by maintaining a stable frequency level, the provision of operating reserve is required and involves additional costs. Apart from conventional generation plants, the storage of electricity or demand-side management are options to provide operating reserves. Countries apply different balancing services resulting in different types and regulations of balancing measures and costs. The European Network of Transmission System Operators for Electricity (ENTSO-E) differentiates three types of reserve (primary,9 secondary10 and tertiary11 control) according to the required speed of response and different reserve needs. Holttinen et al. (2012) provide a categorisation of different reserve types and their application. Costs arising from balancing services for renewable power plants can be assessed if data on forecast errors and the respective cost of the required balancing power are available. Ideally, these values are estimated based on information with high resolution in time (for example, hourly information), since the price of the balancing power depends on the individual load situation. Thus, the use of a modelling tool is suggested in order to estimate the price of the required balancing power in the respective load situation. Depending on the objective, balancing costs for all renewable power plants in a system may be estimated or the difference in balancing costs between a scenario with higher shares of renewable and a reference scenario may be estimated. It should be kept in mind, however, that data availability or estimations of forecast errors (in particular for the future) are subject to uncertainties and may complicate the estimation of balancing costs. Additional profile costs With regard to the required back-up capacity, it should be noted that generation capacities must be provided on a long-term basis to maintain supply security and avoid system break downs and, therefore, to meet projected load or peak loads. Costs resulting from additional or backup capacities needed to provide electricity in times of peak load can also be termed “adequacy costs” (see also Ueckerdt et al. 2013). In an energy-only market, RETs are feeding-in power at nearly zero marginal cost. Consequently, capacity factors of conventional plants decrease as they are less frequently needed, and at least in the short-term,12 average and peak time prices decrease. These effects impair the income situation of all conventional plants, including flexible generation facilities with comparatively high variable costs. Special markets – such as capacity markets – are one option to ensure the long-term provision of flexible capacities such that capital recovery of these plants is ensured. To estimate profile costs for an increased share of RES, the residual load – the system load minus non-dispatchable renewable electricity – should be analysed for extreme situations in the power system such as a high feed-in of fluctuating RES in combination with a high load and low availability of variable RES-E. For the calculation, it is necessary to know the potential gap or the oversupply, respectively, and the frequency of these situations. This can be estimated by means of an electricity market model. To estimate profile costs arising from increased RET use, a reference and an RET scenario are used to compare the resulting costs

Costs and benefits of the energy transition  27

for additionally needed capacity to ensure system adequacy. If no electricity market model is available, generalised results from electricity market models based on the link between capacity factors of installed RETs and the share of RETs in the system could be used to estimate the share of costs arising from RETs. The costs of curtailment can be estimated by multiplying the curtailed electricity by the actual electricity market price as foregone revenue or with the location-specific LCOE as cost. Additional grid costs The use of RES-H generation is typically decentralised and located close to consumption. There is practically no difference in using renewable heat as compared to the use of conventional heating power plants except for solar thermal heat. Further, the existing gas network could be easily used for either natural gas or methane. Similarly to balancing costs, there are hardly any additional grid costs caused by RET-H use. The investment operation expenditures of new or expanded district heating networks are included in the system costs and accounted for in the heat price paid by final consumers. Thus, the additional costs of an expansion of a district heating system are based on the comparison of all expenditures (investment and O&M) at the system level compared to decentralised reference technologies such as gas boilers that are (hypothetically) replaced by the district heating system. In contrast to RES-H, the deployment of RES-E may require the reinforcement and extension of the grid infrastructure. The physical distance between the location of RE power plants in areas with favourable resource conditions and load centres is the major reason for potential bottlenecks in the transmission grid. This mismatch calls for extensions or reinforcements. In addition, there may be bottlenecks in distribution grids, where too much RES-E from smaller power plants such as PV is (irregularly) fed into the grid and, therefore, might endanger the stability of the voltage level and, hence, grid stability at this nodal point. Investments in technical solutions like remote control, interconnectors or higher voltage cables help to overcome these bottlenecks. Additional grid costs should encompass only the investments caused by RET deployment which would not have occurred in an energy system without RE use. The grid extension or reinforcement costs are calculated as the annualised investment over the lifetime of the grid infrastructure and the discounted annual operation and maintenance costs. The detailed estimation of additional grid costs requires a grid model where a scenario with increased RET use is compared to a reference scenario. Additional transaction costs Any economic activity entails transaction costs. Transaction costs accrue in two areas: (i) In the market between actors in the power system (suppliers, demander, grid operator, etc.) (market transaction costs) (ii) At the interface between grid operators and regulatory and administrative bodies (implementation costs) Transaction costs among market participants comprise forecasting, planning and monitoring electricity supply and demand, procuring electricity, establishing markets, contracting, exchanging data, etc. Since these processes tend to become more complex for variable RES-E,

28  Handbook on the economics of renewable energy

transaction costs could increase with RET deployment. As these costs occur mainly at the level of the transmission system operator (TSO), their assessment relies on data available from the TSO. To meet the RE policy targets or requirements, implementation costs arise through monitoring RE activities (such as investments and generation) or reporting requirements or standards in accounting, etc. For example, in Germany, grid operators are obliged to assess the RE levy for final consumers in advance to monitor RE power generation or to update the RE plant register. In addition, the government is obliged to publish monitoring reports on renewable electricity technologies (RET-E) deployment, while in the RES-H sector, it has to monitor and control the RE obligation in new buildings and process, administer and implement special support programs. Therefore, costs occur for grid operators and the government. Figure 2.2 depicts the different types of additional costs and shows the aggregation level at which they could be estimated. If no detailed data are available on additional balancing, grid and transaction costs, etc., and if an energy sector model is available, estimates could be made at the system level. The more detailed data on costs and other relevant items are accessible, the better the quality of the assessed costs. But it should be ensured that no double counting or gaps occur when working at different aggregation levels. In general, it is important to note that interactions between the cost components exist, meaning that the decomposition into the individual components should be seen as an approximation. In this section, we will not analyse the system cost components since the cost effects of the different cost types are very different. Other costs There are further negative effects of an RET-based energy system such as the reduction of fishing grounds, detours or deviation of ship routes, emission of noises and lights/flash, irritations for aviation and radar, killing of birds, etc. Most of these effects are external effects as the actor who is causing them typically does not pay for them. Moreover, some are considered as environmental effects, but they cause negative economic impacts through health impacts or foregone revenues or increased expenditures. As data on these effects are hardly available and it is difficult to account for them in monetary terms, these data are not assessed in the analysis. Benefits from RET use The use of RETs does bring along additional benefits. As mentioned in the German Renewable Energy Law, positive effects are expected regarding climate change and air pollution, fossil fuel savings, decreasing dependency on fuel imports and technological development leading to lower costs of RETs. These effects occur at the system, technology or society level. So far, one major benefit of RET deployment that has been quantified is the reduction of GHG emissions and air pollutants. The benefit of this reduction is a second-order effect such as the alleviation of climate change, health issues, biodiversity, land use and material damages. Since emissions are non-monetary effects arising from the system level, their reduction is considered as a non-monetary benefit at the system level. It can be transformed into a monetary value based on the monetarised value of its second-order effects or on CO2 prices. Avoided emissions of GHG and air pollutants Avoided emissions of greenhouse gases and air pollutants are a major benefit of RET deployment. While GHG emissions have a fundamental impact on climate change and, hence, cause long-lasting global effects, emissions of air pollutants cause more short-term local effects. The

Costs and benefits of the energy transition  29

quantified monetary costs of these emissions are based on the estimated costs of damages that are caused by them or the respective price for CO2 emission allowances. There have been plenty of studies on emission factors for GHG and air pollutants of generation technologies. The more differentiated the emission factors are at the technology level, the better the estimates on emissions. In some countries emission factors for one technology such as coal firing or geothermal heating represent an average value of the existing mix of coal firing technologies or thermal heating technologies, while other studies further differentiate emission factors by the type of coal firing technologies. Direct emission factors reflect the emissions of fuel use (generation) while indirect emission factors represent the emissions related to the provision of the technology or fuel. Ideally, both factors should be included. The monetarisation of emissions in terms of social costs of carbon is very complex (see Tol 2005 for an in-depth discussion). Thus, existing estimations show a large range of values (from €0 to more than €200/t CO2) that are estimated based on diverse models that take into account the effects of GHG on health, consumption, land use, water resources, etc., over a long-term time horizon. Discounting them to the present provides a monetary value for the potential damage of one unit of GHG emitted. Furthermore, discussions arise on whether the marginal damage value of CO2 or the marginal mitigation cost of CO2 should be applied for the monetarisation of CO2 emissions. Under an efficient CO2 certificate market (ETS system), the CO2 price is an alternative option to monetarise the effects. In contrast, damage costs of air pollutants are estimated based on diverse observations, experiments and response functions. They depend to a large extent on where and when they are emitted. Pollutants in densely populated areas cause more damage than in remote areas. Further, the higher the emission location (chimney) the larger the emission zone. As the exact location of the pollution is hardly known, a weighted average of the local damage costs of air pollutants is often used. To assess the avoided emissions at a system level, the amount of power or heat generated per technology and the damage costs should be known. Multiplications of technology-specific emission factors by the amount of power generated by that technology and the damage costs lead to the monetary damage of emissions. The difference between the emission costs of an RE-based and a non-RE-based system shows the benefit – monetary effect – of avoided emissions (damages) at the system level. To assess avoided emissions at the technology level, substitution factors are used that show to what extent fossil energies are replaced by RETs. Technology-specific emission factors indicate the direct and indirect emissions per kWh generated. The quantity of RE generated as well as the damage cost of emissions are further input factors for the assessment. The calculation is depicted in three steps (corresponding to Equations 2.4, 2.5 and 2.6):



Specific avoided emissions per RET and pollutant ( ik )( spVFik )

EF: emission factor in mg/kWh SF: substitution factor i: RE technology j: reference energies k: diverse GHG and air pollutants

= å j ( SFij * EFjk - EFik )

(2.4)

30  Handbook on the economics of renewable energy



Avoided emissions per ik (VEik ) = åi ( spVFik * Qi ) (2.5)

Q: quantity of heat or power generated with RE sources in MWh spVF: specific avoided emission

Avoided social costs per RET i ( soC ) = å k ( VE ik * SK k ) (2.6)

SK: damage costs in euro/t of GHG and air pollutants (damage costs based on different literature sources, or CO2 prices, e.g. ETS) VE: avoided emissions The assessment of the social costs is the most critical step since there is no unique scientifically accepted “damage value”, but a bundle of approaches and model results which provide a wide range of values. Other effects Besides reduced emissions, further positive effects of RET deployment eventually emerge in discussions on the benefits of RET use. Among them, the decrease of RET costs in line with the increasing use of RES (although this is the motivation for RE support) as well as energy security aspects such as decreasing dependence on fuel imports are often discussed. Some of these positive effects such as learning by doing and economies of scale13 are captured via prices in the system-related costs. Nevertheless, some studies strive to measure innovative activities by counting patenting activities as a proxy – but it is difficult to allocate economic values to them and, in addition, innovations are not necessarily reflected in patent applications (since not all innovations lead to a patent). A further positive effect, the spill-over effect of RET technologies on other technologies or other economies, could be captured by patent citations (technology), co-patents and copublications (economy), but a monetary quantification of this effect is not feasible either. Energy security:14 in a functioning market, the scarcity of energy sources is reflected by their market price. This means that any supply constraints – be it through political or economic shocks – typically lead to higher fuel prices. The fuel prices are included in the LCOE, but their volatility and the associated hedging costs might be reflected in these prices to a limited degree. Overall, scarcity increases the generation costs of fossil-based systems (reference systems), and additional generation costs of RET deployment decline under ceteris paribus. In addition, to avoid severe supply constraints, storage of gas or oil serves as insurance against shortages. Adding the insurance costs to fuel prices, generation costs of fossil fuels (reference system) increase and subsequently reduce the additional generation costs of RE use under ceteris paribus. If these costs are not added to fuel costs but are borne, for example, by the public budget, they should be taken into account as separate system-related effects. If stocks become redundant with increasing RET deployment, these avoided stock or insurance costs can be accounted as benefits. Macro-Level Effects of RET Deployment To get an overall picture of the impact of RET use, the effects at the system and actor level should be integrated into a wider overall perspective – the perspective from the macro-economic level. Additional generation costs, grid extensions or surcharges of consumers can

Costs and benefits of the energy transition  31

be measured at the macro-economic level with different macro-economic indicators, such as investments, changes in trade, etc., but typically the overall impact on the economy is expressed in general by changes in GDP or employment. Macro-economic effects show how and to what degree the use of RET affects the economy either in some selected sectors, for example at the RE sector level (sectoral), or in all sectors of the economy (economy wide); this means across all industries and services of an economy. In this context, the terms gross – sectoral – and net – economy-wide – are commonly used. But when talking about gross and net effects, the definition is not always that clear as there is a large variety of macro-economic impact assessments that range between pure sectoral and economy-wide effects. To point out the principal differences between these effects, the following definitions are applied: Sectoral effects: ●

●

●

● ● ●

Account for investments in RETs, operation and maintenance of RETs and depict the activities at the so-called RET sector that could but may not comprise power, heat and mobility. In the impact assessments of the European Commission these expenditures are derived from energy system models. Do not take into account the effects in the conventional energy sector (replacement) and changes in energy prices and income in all other sectors. Depict in some cases direct job effects from RET deployment as well as indirect effects that comprise the upstream industry of the RET sector. Are based on a reference scenario that is derived from energy modelling. Are elements of a sectoral impact analysis that ignores negative effects. Are often called “gross” effects.

Economy-wide effects: ●

● ●

●

Take into account all positive and negative effects (direct, indirect, induced) including changes in prices and rents that in turn affect consumption, production and income. Include impacts in all sectors of the economy, along the value chain. Rely on a comparison of two different RET deployment scenarios (no or low RET vs advanced RET use) that are derived from macro-economic models. Are often called “net” impacts of RET use.

RE sector effects Gross effects show the employment in RE industries and service sectors such as RET manufacturing, or RE project planning, operation of an RE generation facility, etc., and in its related industries. In principle, these effects are assessed only for RET-based energy systems and do not rely on a comparison of two systems (with and without RET use).15 So-called gross impact assessments are related only to industries and service sectors directly involved in RE activities, and only look at positive effects in these industries. They are also called sectoral effects at the macro-level. Consequently, gross impact studies are sectoral studies depicting the significance and relevance (share of employment) of RETs in an economy. Several indicators are commonly used to illustrate the RET deployment-induced effects in the RE sector. They comprise investments in RETs and turnover of RET manufacturers in

32  Handbook on the economics of renewable energy

the respective sector, avoided imports, jobs in the RE sector (plus upstream industry), value added, etc. Investment and turnover Investments in RETs are used as a common macro-economic indicator to highlight the significance of RETs for economic activities. At the macro-economic level, investments in renewable energy generation technologies are the main impulses triggering further economic activities such as manufacturing of RETs or intermediate products for RETs. The economic activities that are triggered by investments can be measured as the turnover of RE-related manufacturing and service sectors. This shows the demand for RETs, services, equipment or components in the respective manufacturing industries. To assess the total (new) investments in RETs, two approaches are feasible: a) An assessment of investment expenditures per installed capacity as well as the newly installed capacity per technology is needed. However, investments in RETs or services cannot be translated completely into domestic effects, because part of the technology or services could be imported and not all domestically produced technology components are domestically installed. In order to use investment expenditures as an impulse for economic growth (investment impulse), the national RET investment expenditures have to be corrected by the net imports.16 The calculation is depicted in Equation 2.7.

Investment impulseRET * = specific investment expendituresRET * installed capacityRET - net importsRET

(2.7)

Investment impulses are then split into renewable energy technology components and services and are allocated to the respective sectors. The result discloses the turnover of manufacturers and service providers in these sectors. Turnover from manufacturing is assessed on the basis of newly installed capacities minus imports plus exports. Turnover from operation and maintenance is assessed on the basis of generation or cumulated installed capacities. Taking exports into account is especially important for countries with high export shares. b) Alternatively, investments could be derived from energy sectors’ models which could include import options as well. The model results of the RET and reference scenario indicate which investments are due to RET deployment. Sectoral employment (RE sector) Sectoral or so-called gross employment shows how many jobs (in full-time equivalents or “fte”) exist in the sectors that are involved in any RE activities such as manufacturing, project development, research, operation, etc. It reveals the significance of RE sector employment in comparison to total employment (see also Chapters 4, 5 and 6). Furthermore, if sectoral employment is broken down to the technology level, the importance and dependence of an economy on the respective technology can be shown. Gross employment can be assessed by means of different approaches. A very common approach is the use of employment factors, which relies on labour coefficients indicating fulltime equivalent jobs needed per MW installed (manufacturing, construction, installation) or

Costs and benefits of the energy transition  33

per MWh generated (operation and maintenance). Simple multiplication of these coefficients with additionally installed capacities and generation, and some adjustments for regional or technological factors and ex/imports lead to gross employment figures that show the number of direct jobs. A more complex assessment approach uses, for example, input-output tables (see Chapter 5). In this approach, investment expenditures (export/import adjusted) of RETs are divided into components and cost shares, which are in turn allocated to the respective economic sectors. They are used as impulse (input) in the input-output model to assess changes in industry production and services due to RET use. A multiplication of the output changes by labour coefficients displays the number of jobs created due to RET deployment. Imports or exports of fossil fuels and technology Avoided imports of fossil fuels due to RET deployment reflect a macro-economic figure that highlights the decrease of import dependency and geopolitical risks of energy supply. It should be noticed, however, that a decrease of imports might not necessarily entail a positive economic impact per se. If imports are replaced by products that are “more expensive” per unit of energy, consumers might face a loss in their consumer surplus. It is difficult to translate the value of a decreased import dependency into monetary terms. One option is to look at the risk of price fluctuations and shortages of certain fuels (geopolitical risks). Both might decrease in line with decreasing fossil fuel imports. The reduced risk may lead to lower requirements for storage capacity for gas, for example, and thus reduce infrastructure-related costs. These reduced “security” costs could be counted as a “benefit” at the system level, but again their monetary quantification is a challenging task (see Lehr 2011). The economic effects of increases in technology exports are captured by or reflected in production, turn-over, value added or gross employment data. Subsequently, the trade or export of RETs is a macro-economic indicator, which enters macro-economic modelling and might have a large impact on value added. But the exports themselves do not allow any conclusions on the actual “net” benefits. Net effects At the macro-level, the economy and society as a whole are affected. The impacts comprise economic activities and welfare, wellbeing, health and the environment. Typically applied assessment criteria of net effects at the macro-level, for example in the impact assessments of the European Commission, are: ●

●

●

GDP: this depicts how much welfare a country’s economy gains or gives up due to RET deployment. It is an economic indicator that stands for welfare. Net employment: this displays economy-wide impacts on the number of jobs due to increase in RE deployment. The effect is depicted as a cumulative term – additional jobs over the total period – or as an annual term – the number of jobs in the year(s) of consideration. Investment activities entail temporary employment, while operation and maintenance create a small but long-lasting employment effect. Health effects: these are effects on the health of people in society and are expressed in monetary values, i.e., as avoided health expenditures. We can apply this as an indicator for wellbeing, but it is limited in this regard, since it does not capture all the effects on wellbeing.

34  Handbook on the economics of renewable energy ●

Avoided emissions: we account for these at the system level, but when depicting the consequences of CO2 emissions in terms of losses in GDP or overall consumption, it is a macro-level effect and could be endogenous.

The assessment of net effects implies the development of two scenarios; this means a RET and a reference scenario. All positive as well as negative effects of RET deployment should be included (see Table 2.3) to get a real overall picture of the RET effects. Ideally, all costs and benefits at the system-level as well as the different charges and reliefs of economic agents at the micro-level should be included when assessing net effects at the macro-level.

CONCLUSIONS As the energy transition progresses in the EU and other parts of the world, identifying the nature and estimating the magnitude of associated positive and negative effects, i.e. costs and benefits, remains a major challenge. This chapter provides guidance in this respect by developing a consistent analytical approach. This includes an introduction of the necessary terminology and an assessment of different types of costs and benefits of RET deployment. Conceptual Structure of Costs and Benefits An analytical framework helps to establish the research question and identify its boundaries by highlighting on which impact levels costs and benefits occur. In addition, we distinguish between different possible perspectives for the assessment. In the first step, it is necessary to establish the boundaries of the system that one is seeking to assess. This entails defining the Table 2.3  Overview of positive and negative effects that should be taken into account when modelling net effects of RET deployment Positive effects → job increases

Negative effects → job losses

Increase in investment in RETs (RE industry and upstream industry)

Displaced investment in conventional generation technology (CE industry and upstream industry)

Increase in O&M in RE generation (RE industry Displaced O&M in conventional power generation and upstream industry) (CE industry and upstream industry) Increase in fuel demand (biomass) (RE industry and upstream industry)

Decrease in demand for fossil fuels (CE industry and upstream industry)

Increase in trade of RE technology and fuel (biomass) (RE industry and upstream industry)

Decrease in trade of conventional technology and fossil fuels (CE industry and upstream industry)

Higher household income from employment in the RE industry

Lower household income from employment in CE industry

Decreased electricity price for households and industry due to merit-order effect, CO2 pricing, etc.

Increased electricity price for households (budget effect) and industry (cost effect) due to additional generation cost of RE-based power generation

Source:   Breitschopf et al. (2013)

Costs and benefits of the energy transition  35

focus of research, i.e. the research object, the perspective or impact level that one will take into account (macro-, meso- and micro-level), the character of the impact analysis (economic, environmental, technological, social and/or societal), potential interactions with other systems or sectors as well as the geographic area and time horizon at hand. It will be equally important to select an appropriate counterfactual situation, i.e. a reference scenario against which the research object will be compared. There are several possible approaches that one might apply. This includes before/after comparisons, control groups, reference technologies, trends, minimum standards, current shares or target achievements as well as modelling results. When referring to modelling results, reference scenarios can, e.g., be based on assumed developments of the energy system without RETs or support policies. Cost and Benefit Categories: The Case of Renewables Overall, effects of the energy transition can occur at three different levels. First, there are actor-specific effects considering a rather micro-economic perspective. The second category refers to system-related effects, which encompass all direct and indirect costs and benefits of the energy transition within a defined system compared to the reference system. Third, macrolevel effects reflect effects in an economy. In addition to their primary consequences, effects may also have so-called second-order effects, e.g. air pollutants have a detrimental effect on health which, in turn, could influence income. Actor-specific effects of RET deployment The first category we focus on is actor-specific effects of RET deployment. These are essentially distributional effects from the actor perspective, e.g. firms, households or the public sector. Deploying RETs may lead to changes in costs, prices, revenues, quantity or quality of services and products. Whilst benefits might accrue to specific actor groups, others may be negatively impacted by the same intervention. Overall, we can distinguish direct effects, indirect price effects and non-monetary effects. Direct effects usually constitute financial effects such as direct payments, levies, remuneration schemes or investment grants. Indirect price effects arise through changes in the energy price or technology price as a result of the market mechanism, i.e., these involve changes in the quality of goods, their supply and demand. Examples include capitalisation and scarcity rents as well as a higher consumer surplus through declining electricity prices due to the increasing share of RETs in generation. Finally, non-monetary effects comprise impacts on the individual utilities that arise due to non-monetary changes. They have to be expressed via proxies. Following Breitschopf and Diekmann (2013), we can further categorise actor-specific effects into six types:

1. Effects on consumer surplus 2. Effects on producer surplus 3. Scarcity rents 4. Capitalisation effects 5. Changes in utility 6. Transitional effects and others

In both the power and heat sectors, the deployment of RETs has been supported by a variety of policy instruments ranging from price- or cost-based support to quantity-based support. If

36  Handbook on the economics of renewable energy

the use of RETs causes additional costs at the system level, these costs have to be borne by someone. The financing of these costs is determined by regulations. In principle, the financing of RET deployment can rely on two basic financing options that can be combined as well: 1. Consumer-based financing: in the case of consumer-based financing, the deployment of RETs is financed by final consumers and no public budgets are used. Examples are feedin premiums or quota systems. 2. Budget-based financing: as for budget-based financing, this means the financing of RET deployment by government or state budgets. The budget can be used to finance feed-in premiums, investment grants or tax credits for RE generation or capacities. System-related effects of RET deployment The second category of effects of RET deployment is system-related effects. Here, the term “system” may either refer to the energy sector as a whole, a final energy sector or it may be broken down to the technology level. System-related effects comprise the costs and benefits that arise from integrating RETs into the energy system. They are assessed without taking into account taxes or subsidies or other policy-induced transfer payments for investment or generation. In general, costs and benefits occurring in a system with significant shares of RETs (the “RET-scenario”) and those of a reference system (the reference, business-as-usual or baseline scenario) are compared against each other. It is important to exclude double-counting. Additional system costs can roughly be differentiated into two types of costs, additional direct and indirect system-related costs. Whilst additional direct costs can be assessed at the technology level and reflect only the costs arising from the generation of heat or power with RETs minus the avoided costs of conventional, fossil-fuel-based generation, additional indirect costs may also include costs that are not directly linked to electricity generation. In the case of the electricity sector, they usually occur with regard to integrating RETs into the system and ensuring a stable power system. Indirect system-related costs can be grouped into balancing costs, profile costs, grid costs and transaction costs. Balancing costs arise from the fact that most RETs are intermittent and variable in nature and thus not fully dispatchable. Forecasts errors will have to be “balanced”, so that the total supply is equal to the demand profile. Profile costs reflect the need for medium-/long-term flexibility and refer to the back-up capacity needed to ensure system and supply security. Grid-related costs refer to reinforcements or extensions of transmission or distribution grids as well as congestion management including the redispatch which is required to manage situations of high grid loads in the electricity sector. Lastly, two types of transaction costs will also occur, on the one hand costs between actors in the power system (referred to as “market transaction costs”), and on the other hand costs at the interface between grid operators and regulatory and administrative bodies (referred to as “implementation costs”). In addition, other costs which are often hard to monetise might accrue, most often in the form of external effects, e.g. interference with bird routes or fish breeding grounds, emission of noise or light pollution, etc. Additional system benefits can be expected with regard to climate change and air pollution, fossil fuel savings, decreasing dependency on fuel imports and technological development. The major benefit of RET deployment is avoiding emissions of greenhouse gases and

Costs and benefits of the energy transition  37

air pollutants. The quantified monetary costs of these emissions are based on the estimated costs of damages that are caused by them or the respective price for CO2 emission allowances. In contrast, damage costs of air pollutants are estimated based on diverse observations, experiments and response functions and depend to a large extent on where and when they are emitted. Besides reduced emissions, further positive effects of RET deployment might emerge including the decrease of RET costs in line with the increasing use of renewable energy sources as well as energy security aspects such as decreasing dependence on fuel imports. Macro-level effects of RET deployment Third, macro-level effects of RET deployment bring together and integrate the effects occurring at the system and actor level. Additional generation costs, grid extensions or surcharges of consumers can be measured with different macro-economic indicators, such as investments, changes in trade, etc., but typically the overall impact on the economy is expressed by changes in GDP or employment. In general, one can distinguish between economy-wide effects (usually referred to as “net” impacts of RET use) and sectoral effects (usually called “gross” effects). Sectoral effects account for investments in RETs, operation and maintenance of RETs and depict the activities at the so-called RET sector that could but may not comprise power, heat and mobility. Several indicators are used to illustrate the RET deployment-induced effects in the RE sector. They comprise investments in RETs and turnover of RET manufacturers in the respective sector, avoided imports of fossil fuels due to RET deployment (i.e. decrease of import dependency), sectoral employment in the RE sector (plus upstream industry), value added, etc. Economy-wide effects take into account all positive and negative effects (direct, indirect and induced) including changes in prices and rents that in turn affect consumption, production and income and comprise impacts in all sectors of the economy, along the value chain. The impacts include economic activities and welfare, wellbeing, health and the environment and are assessed by comparing two scenarios, a RET and a reference scenario. Assessment criteria for net effects are GDP, net employment, health effects and avoided emissions. Areas for further research As the energy transition progresses and new geographical areas and less mature technologies (e.g. offshore wind, wave and tidal energy or fourth generation district heating systems) are increasingly being deployed, the assessment of costs and benefits will continuously have to evolve to accommodate new effects. Furthermore, as the overall shares of RES are increasing in the EU energy system, additional balancing, profile and grid-related costs will likely gain in significance and robust modelling approaches will have to be developed. Another strand of research relates to assessing the costs and benefits of so-called “hybrid projects” which combine generation and infrastructure assets. Here, it will be necessary to bring together existing approaches and methodologies. Additionally, it will be crucial to pay increased attention to the social effects of increased RET deployment and distributional effects on different societal groups. Furthermore, the choice of an adequate reference system or counterfactual could be reconsidered, provided that a development without RETs is not a real option anymore. Applying and adapting this concept to other areas such as energy

38  Handbook on the economics of renewable energy

efficiency measures or decarbonisation options in industry or transport is another topic for further research.

NOTES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

11. 12. 13.

14. 15.

16.

https://www​.impres​-projekt​.de​/impres​-en/. For a description of economic cost benefit analysis (CBA) applied to environmental topics we refer to Hanley and Spash (1993). https://www​.impres​-projekt​.de​/impres​-en/. Breitschopf and Held (2014). Breitschopf and Billerbeck (2021). See https://publica​.fraunhofer​.de​/eprints​/urn​_nbn​_de​_0011​-n​-6384598​.pdf for an overview, or https://epatee​.eu​/ knowledge​-base. Tax credits or subsidies can be granted for own generation and consumption as well as for the consumption of third parties. The technology mix reflects a mix of different conventional technologies that are replaced by one single RET. Primary reserves (or frequency containment reserves) designate the active power reserves available to contain system frequency after the occurrence of an imbalance (ENTSO-E Glossary). Secondary reserves (or frequency restoration reserves) designate the active power reserves available to restore system frequency to the nominal frequency and, for a synchronous area consisting of more than one load-frequency control area, to restore power balance to the scheduled value (ENTSO-E Glossary). Tertiary reserves (or replacement reserves) designate the active power reserves available to restore or support the required level of secondary reserves to be prepared for additional system imbalances, including generation reserves (ENTSO-E Glossary). In the mid- to long term, the conventional power plant should adapt to the new circumstances: base load plants are expected to become less profitable with increasing shares of renewables, and thus they will be replaced by more flexible technologies. Learning by doing refers to learning from working experience, increasing the skill in production and exchanging knowledge within the firm (Alhusen et al. 2021). Economies of scale refer to the pure size effect – the larger production capacities, the lower the fixed costs of a unit of output (lower unitary fixed costs). Economies of scale do not reflect increases in efficiency. Security refers here to fuel supply (such as import dependency) and not to the reliability of the power system. The latter is taken into account by balancing and grid costs. If employment levels of with-RETs and without-RETs energy systems are compared without taking into account potential price and income effects of RETs in other sectors, then two gross effects (sectoral effects) are compared with each other without depicting their impact on the whole economy. Single parts or components are imported, which represent a share of investment expenditures for a RET.

REFERENCES Alhusen, H., Bennat, T., Bizer, K., Cantner, U., Horstmann, E., Kalthaus, M., Proeger, T., Sternberg, R., and Töpfer, S. (2021). New measurement conception for the ‘doing-using-interacting’ mode of innovation. Research Policy 50: 1–15. Breitschopf, B., and Billerbeck, A. (2021). Report on results of the meta-study: Factors governing decisions in H&C. http://publica​.fraunhofer​.de​/dokumente​/ N​-638442​.html. Breitschopf, B., and Diekmann, J. (2010). Vermeidung externer Kosten durch Erneuerbare Energien – Methodischer Ansatz und Schätzung für 2009 (MEEEK). In Rahmen des Projekts “Einzel- und

Costs and benefits of the energy transition  39

gesamtwirtschaftliche Analyse von Kosten- und Nutzenwirkungen des Ausbaus Erneuerbarer Energien im deutschen Strom- und Wärmemarkt,” June 2010 Untersuchung im Auftrag des Bundesministeriums für Umwelt, Naturschutz und Reaktorsicherheit. https://publica-rest.fraunhofer. de/server/api/core/bitstreams/c3daf333-e4d7-41f5-97b8-65db83cecd23/content. Breitschopf, B., and Diekmann, J. (2011). Gesamtwirtschaftliche Auswirkungen des Ausbaus Erneuerbarer Energien. In Markus Gerhard, Thomas Rüschen, and Armin Sandhövel (Eds.), Finanzierung Erneuerbarer Energien. Frankfurt School Verlag, October 2011. Breitschopf, B., and Diekmann, J. (2013). Verteilungswirkungen erneuerbarer Energien – Grundlagen, Systematik und methodische Ansätze zur Erfassung (Distributional effects of RE – fundamentals, systematic and methodological approaches). Internal working paper. http://www​.impres​-projekt​.de​/ impres​-de​/content​/veroeffentlichungen​.php. Breitschopf, B., and Held, A. (2014). Guidelines for assessing costs and benefits of RET deployment: DiaCore, D4.1, 42 pp. https://www.isi.fraunhofer.de/content/dam/isi/dokumente/ccx/dia-core/D4-1_ Guidelines_for_assessing_costs_and_benefits_of_RET_deployment.pdf. Breitschopf, B., and Winkler, J. (2019). The EU 2030 renewable energy vision: Can it be more ambitious? Advances in Environmental Studies 3: 164–178. Breitschopf, B., Nathani, C., and Resch, G. (2013). Employment impact assessment studies: Is there a best approach to assess employment impacts of RET deployment? Renewable Energy Law and Policy Review 2: 93–104. Breitschopf, B., Held, A., and Resch, G. (2016). A concept to assess the costs and benefits of renewable energy use and distributional effects among actors: The example of Germany. Energy & Environment 27: 55–81. Diekmann, J., Schill, W.-P., Breitschopf, B., Sievers, L., Klobasa, M., Lehr, U., and Horst, J. (2016). Impacts of renewable energy deployment – Summary and conclusions: Final report in the framework of the project “Impacts of Renewable Energy Sources (ImpRES)” supported by the Federal Ministry of Economics and Energy. Deutsches Institut für Wirtschaftsforschung e. V; Fraunhofer-Institut für System- und Innovationsforschung; Gesellschaft für Wirtschaftliche Strukturforschung mbH; Institut für ZukunftsEnergieSysteme. http://www​.impres​-projekt​.de​/impres​-en​/content​/abschlussworkshop​.php. ENTSO-E. (2022). Glossary. Definitions and Abbreviations. https://www​.entsoe​.eu​/outlooks​/midterm​ /glossary/. Fullerton, D. (2009). Distributional effects of environmental and energy policy: An introduction. In: Fullerton, D. (Ed.), Distributional Effects of Environmental and Energy Policy. London: Routledge, pp. xi–xxvii. Hanley, Nick, and Spash, Clive L. (1993). Cost-Benefit Analysis and the Environment. Aldershot, Hants, England, Brookfield, Vt: Edward Elgar Publishing. Holttinen, H., Milligan, M., Ela, E., Menemenlis, N., Dobschinski, J., Rawn, B., Bessa, R. J., Flynn, D., Gomez-Lazaro, E., and Detlefsen, N. K. (2012). Methodologies to determine operating reserves due to increased wind power. IEEE Transactions on Sustainable Energy 3(4): 713–23. ISI, GWS, IZES, DIW. (2010). Einzel- und gesamtwirtschaftliche Analyse der Kos-ten- und Nutzenwirkungen des Ausbaus Erneuerbarer Energien im deut-schen Strom- und Wärmemarkt. Barbara Breitschopf, Marian Klobasa, Frank Sensfuß, Jan Steinbach, Mario Ragwitz, Ulrike Lehr, Juri Horst, Uwe Leprich, Eva Hauser, Jo-chen Diekmann, Frauke Braun, Manfred Horn. Studie im Auftrag des Bundesministeriums für Umwelt, Naturschutz und Reaktorsicherheit. Zwischenbericht, March 2010. http://www​.erneuerbare​-energien​.de​/unser​-service​/mediathek​/downloads​/detailansicht​ /artikel ​/ kosten​-und​-nutzenwirkungen​-des​-ausbaus​-erneuerbarer​-energien​-im​-deutschen​-strom​-und​ -waermemarkt/​?tx​_ttnews​%5BbackPid​%5D​=98​&cHash​=985​5906​4b8b​7e8f​6946​4b75​1ab551a87. ISI, GWS, IZES, DIW. (2013). Monitoring der Kosten- und Nutzenwirkungen des Ausbaus Erneuerbarer Energien (Monitoring of costs and benefits of RET deployment), Studie im Auftrag des Bundesministeriums für Umwelt, Naturschutz und Reaktorsicherheit, Update für 2012, September 2013. http://www​.impres​-projekt​.de​/impres​-de​/content​/veroeffentlichungen​.php. . Lehr, U. (2011). Methodenüberblick zu Ermittlung vermiedener Brennstoffimporte. http://www​ er neuerba re ​ - energien ​ . de ​ / unser ​ - ser vice ​ / mediat hek ​ / down loads ​ / det a ila nsicht ​ / a r ti kel ​ / methodenueberblick​-zur​-abschaetzung​-der​-veraenderungen​-von​-energieimporten​-durch​-den​-ausbau​ -erneuerbarer​- energien/​ ? tx ​ _ ttnews​ %5BbackPid​ %5D ​ = 98​ & cHash ​ = e83​ f ff4​ 2 d4f ​ 6 d86 ​ b945​106d​ c70b2cf94.

40  Handbook on the economics of renewable energy

Lutz, C., and Breitschopf, B. (2016). Systematisierung der gesamtwirtschaftlichen Effekte und Verteilungswirkungen der Energiewende (Systemic overview of macroeconomic and distributional effects of the energy transition). GWS Research Report 2016/1, 56 pp. http://papers​.gws​-os​.com​/gws​researchreport16​-1​.pdf (accessed 9 July 2018). Sievers, L., Breitschopf, B., Pfaff, M., and Schaffer, A. (2019). Macroeconomic impact of the German energy transition and its distribution by sectors and regions. Ecological Economics 160: 191–204. Tol, R. S. J. (2005). The marginal damage costs of carbon dioxide emissions: An assessment of the uncertainties. Energy Policy 33: 2064–2074. Ueckerdt, F., Hirth, L., Luderer, G., and Edenhofer, O. (2013). System LCOE: What are the costs of variable renewables? Energy 63: S61–S75. http://www​.sciencedirect​.com​/science​/article​/pii​/ S0360544213009390.

3. Energy system modelling of renewable energy and related energy infrastructure needs Gustav Resch, Franziska Schöniger, Florian Hasengst, Demet Suna, Gerhard Totschnig and Frank Sensfuß

INTRODUCTION Energy is a driving element of global society, and it is the main source of global greenhouse gas (GHG) emissions, being responsible for about three-quarters of those emissions today (IEA 2022a). The envisaged transition of the energy system and the whole economy towards decarbonisation poses several questions for policy makers, stakeholders and the various actors from the business and private sectors. Profound analysis is required to support decision making at various perspectives, be it to inform on pros and cons of certain policy directions, to assist in directing investments or to clarify operational aspects of available energy system assets. Here is where energy system modelling comes into play, gaining renewed attention in the public and the scientific debate. Thanks to advancements in computational capabilities, an increase in the number of energy system models is observable over the last decades. To assist in identifying the right tool for the specific challenge and application, several reviews of energy system models can be found in the literature. Many of them provide a classification of the models assessed, informing on the scope, the underlying methodologies or the challenges faced in modelling. Such classifications may help in identifying differences and similarities of the energy system models and, thus, assist in the process of selecting the proper one (Van Beeck 1999). In recent years, as a key option for decarbonising energy supply, renewable energy (RE) has become a focus in energy system modelling as well as in real-word energy sector developments. This chapter aims to shed light on the issues raised above, showcasing past developments in energy system modelling of renewable energy and related energy infrastructure needs. Via a few snapshots, it provides a chronological survey of energy system models with a focus on renewable energy and the main related thematic fields. Geographically, these snapshots refer to Europe and illustrate topics of interest as well as approaches used in corresponding modelbased assessments. Results gained from these analyses are thereby also summarised in brief.

EVOLUTION OF KEY POLICY QUESTIONS AND RELATED MODELLING ACTIVITIES During the first decade of this century (2000 to 2010) renewable energy stepped out of its niche in Europe as well as globally. The need for combating climate change has dominated energy policy debates and, as a key option for decarbonisation, renewable energy has been prominent in the policy debate, specifically in Europe but also globally. 41

42  Handbook on the economics of renewable energy

Within Europe, apart from large hydropower plants that had mostly been built decades ago, only minor advancements had taken place. Countries like Denmark, Germany and Spain had acted as pioneers for wind power developments during the late 1990s, and slowly other countries have followed their role model of supporting an enhanced uptake of wind as well as other renewable energy technologies. Policy attention has focussed on the electricity sector whereas in the previous century the majority of deployment had taken place in the heat sector due to the predominant share of biomass in heating in several countries. With the ongoing liberalisation of power markets within Europe, the pressure had grown to let independent power producers, largely relying on RE technologies, enter the scene. At that point in time, the large majority of RE technologies were characterised by higher costs compared to the predominant fossil fuel-based generation options due to the low prices for coal and gas at the beginning of this century. So-called “new renewables” like wind power, photovoltaics, concentrated solar thermal power and geothermal energy as well as biomass and biogas thus required financial support to enhance their uptake and to make investments in these technologies economically viable. In accordance with the above, at that point in time key policy topics were: ●

● ●

●

To indicate the prospects for an enhanced market uptake of renewable energy and to assess the feasibility of policy targets for RE in the 2010 and 2020 timeframe. To estimate the level of dedicated RE support necessary to encourage RE market uptake. To analyse the macro-economic impacts of ambitious RE policy targets and energy transformation pathways. To assess the pros and cons of different RE support instruments concerning effectiveness and economic efficiency.

Energy modelling with a focus on renewable energy consequently aimed to provide answers to the issues raised above. A few examples of the types of energy system models that have taken a central role in policy making within the EU are: ●

●

●

●

The PRIMES modelling system, developed and operated by National Technical University of Athens (NTUA), is an energy modelling system that covers all energy sectors, providing a complete depiction of energy supply and demand within Europe. It has been serving as the key modelling tool for in-depth analyses of general energy policy decisions at EU level and has been applied in various impact assessments of the European Commission related to energy and climate policy options. However, at that point in time, PRIMES offered a less detailed representation of RE technologies and related potentials. The Green-X model is a specialised energy system model developed by Technische Universität Wien (TU Wien) that builds on a detailed representation of the potentials and cost of RE technologies within European countries and that allows for assessing the pros and cons of the design elements of different RE support instruments. The Invert/EE-Lab model is another energy system model developed by TU Wien. The model is constrained to the EU building sector, and it allows for an assessment of the interplay between energy efficiency and (renewable) energy supply for the building sector thanks to a detailed representation of the building stock across the EU. The NEMESIS model, developed and operated by SEURECO, is a representative of macro-economic models and has been used to assess wider economic impacts of an enhanced uptake of RE technologies within Europe.

Energy system modelling of renewable energy  43

Over the past decade (2010 to 2020), renewable energy has become mainstream within Europe and also globally. As stated by the International Energy Agency (IEA) (cf. IEA 2022b) and by the International Renewable Energy Agency (IRENA) (cf. IRENA 2022), the majority of global investments in electricity generation assets was directed towards RE technologies. Within Europe, enforcing and enhancing market integration became one of the predominant topics in energy policy making related to renewables, cf. Resch et al. (2013) or Del Rio et al. (2017). In later years, towards the end of the past decade, the long-term perspective for the energy sector transformation also gained increasing policy importance, with an impact on research and related modelling activities. Driven by the Paris Agreement on effectively limiting the global temperature rise to well below 2°C, the EU and its member states agreed on the long-term target of achieving carbon neutrality by 2050 and, consequently, aimed for envisioning pathways for the decarbonisation of the energy sector and the whole economy. In accordance with the above, an incomplete list of other key policy and research topics at that point in time includes: ●

●

●

●

●

●

Adapting the design of RE support schemes to better cope with the rapid decrease in technology costs. Removing so-called non-financial barriers that hinder the uptake of renewables, including improving or simplifying permission procedures, grid access, etc. Improving financing conditions for renewables so that the high up-front costs of RE technologies will be less of a hurdle. Operational and supply security aspects of electricity networks to better cope with high shares of renewables in the future. Infrastructural prerequisites for enhancing the uptake of renewables in the electricity sector and in other sectors. Aspects of energy sector coupling to facilitate the decarbonisation of industry, heat and transport.

Energy modelling gained increasing attention in accordance with the above. That involved, for example, facilitating mid- to long-term energy planning as well as clarifying operational aspects for the electricity sector that come along with sector coupling and with increasing shares of variable renewables like wind and solar. In addition to the models listed above, energy system models with a focus on power plant dispatch in the electricity sector were widely used for topical analyses at that point in time. With improved computational capabilities, a particular spotlight has been put on the power sector and the modelling of the interplay between supply, storage and demand at a high temporal resolution. As stated above, aspects of system integration of variable renewables like wind, solar and run-of-river hydropower, all depending on changing weather conditions, became key from a topical viewpoint thanks to increasing renewable shares. Additionally, transparency in modelling became a prerequisite and, similar to climate modelling, a move towards opensource modelling was observable. A few examples of energy system models with a focus on the electricity sector are: ●

The Baltic Model for Regional Electricity Liberalisation (Balmorel) model, originally developed by the Danish Technical University (DTU), is an example of an open-source, partial equilibrium model for the electricity and combined heat and power sectors. The

44  Handbook on the economics of renewable energy

●

●

main strength of this model is its high modularity, enabling analyses at various geographical levels (regional, national, international) and different additional sectors like electric mobility, individual heating and hydrogen production. The model was one of the first open-source energy system models in Europe as, e.g., documented by Ravn et al. (2001). The Enertile model, developed by the Fraunhofer Institute for Systems and Innovation Research (Fraunhofer ISI), is an energy system optimisation model representing the electricity, grid-connected heat and hydrogen sectors in high temporal, spatial and technoeconomic resolution. Enertile optimises, in an integrated manner, the dispatch of and investment in the supply technology options for the generation, conversion, transmission and distribution of the described sectors. A strength of Enertile is its integrated calculation method of renewable energy potentials and generation profiles for Europe and North Africa with a very high level of granularity which considers detailed regional weather input data and potential limitations like distance regulations and protected areas. The High Resolution Power System (HiREPS) model, developed by TU Wien, is another example of an energy system model which focuses on the electricity sector and has been frequently used in the European context. HiREPS is a power simulation and optimisation model able to dynamically model the electricity and district heating sectors in hourly resolution. The strength of HiREPS lies in its high resolution of hydropower modelling of the Austrian and German electricity sectors, including a high-detail representation of cascades and temporal correlations between different hydropower units. Similarly to other models, the implementation of sector coupling in the form of, e.g., electric mobility including different charging strategies, power to heat and hydrolysis is also captured.

Many of the electricity sector models allow for a brief representation of infrastructural prerequisites, specifically with respect to the cross-border transmission grid. However, a detailed grid analysis requires a wide range of data available at a high level of geographical granularity, including information on generation assets and demand structures, in addition to information on the electricity grid. One model example for the electricity grid infrastructure within Europe is the TEPES model, developed and applied by Comillas University of Madrid (Spain). Another example of models dedicated to identifying energy-related infrastructural prerequisites is the Hotmaps modelling tool which offers heat maps for analysing opportunities for grid-connected heating and cooling supply in the European context. In recent years, driven by increasing awareness of climate change impacts as a consequence of climate-driven phenomena like heat waves, droughts and flooding, a systematic incorporation of expected future climate impacts is gaining attention also in energy system modelling. That may allow for a check-up of the resilience of today’s and the future energy system, in order to safeguard energy supply in times of climate change and with more frequent climatedriven extreme events. Now, the first pilot studies are emerging, but a comprehensive systemic analysis is missing, at least in the European context.

SPOTLIGHT ON SELECTED ENERGY SYSTEM MODELS WITH A FOCUS ON RENEWABLE ENERGY This section is dedicated to approaches taken in energy system modelling with the given topical focus on renewable energy and infrastructural prerequisites. More precisely, via a few

Energy system modelling of renewable energy  45

snapshots, we shed light on examples of energy system models that have put a focus on renewable energy. This is done by describing the underlying modelling approach in further detail, complementary to the brief description provided in the previous section. The models are presented in alphabetical order. Please note that, in the follow-up section, we then showcase the models’ application, providing information on the approach taken and the results derived in exemplarily selected cases of model uses. As a starting point, we provide a general classification of energy system models in accordance with the literature. General Classification of Energy Models Analytical tools such as energy models have emerged as a useful methodology for energy research for example aimed at evaluating future energy supply options and generating insights into some of the associated uncertainties. Today, there are many types of energy models covering a wide range of analytical approaches, with tools often developed for specific objectives and with specific needs, with a predefined methodological scope and limited application. As illustrated in Figure 3.1, in accordance with the literature (cf. Kannan and Turton 2011), we can categorise energy-related models into the following groups: ● ● ●

Macroeconomic models. Energy system models. Sector-specific energy (system) models, including electricity models as well as models that cover other energy sectors.

Source:   own elaboration based on Kannan and Turton (2011)

Figure 3.1  General classification of energy-related models

46  Handbook on the economics of renewable energy

Further distinctions can be made concerning the model’s overall focal area within the energy system, i.e. distinguishing between supply-side and demand-side models. Following the above classification of energy models, the term “energy system models” comprises all sorts of energy models as long as cross-sectoral interactions are incorporated into the underlying modelling approach. Some of the models have a high level of technological detail, while others place a stronger focus on the representation of the energy industry. As broadly illustrated in Figure 3.1, the objectives and scope of these models are diverse, with different strengths and weaknesses that provide complementary insights into a range of aspects of the energy system. One of the key attributes that characterise the application of an energy model is its level of temporal detail. According to Boyd (2016), the temporal representation has three dimensions: (1) the model time horizon; (2) the length of each period; and (3) dissolution within one year. The time horizon is crucial when research is concerned with long-term utility and infrastructure developments needed to address long-term energy challenges (such as climate change or oil shortages). In contrast, a high degree of interannual time resolution is very important when the energy system has to accommodate fluctuating demand and supply of energy commodities, as is the case with high shares of wind and solar in the electricity supply. Ideally, models of energy system evolution should combine a sufficiently long time horizon and adequate intertemporal resolution for the given analysis. In practice, the trade-offs between these two time dimensions are determined by computational constraints, data availability, the type of time-dependent variable, methodological limitations within the modelling framework, etc. Balmorel: An Open-Source Energy System Model for the Electricity and District Heating Sector Open-source energy system models facilitate the transparent dissemination of relevant technoeconomic and policy research analyses. In recent years, open-source modelling has become a new sort of standard in energy system modelling undertaken in research projects at the European level. As an example of an open-source, partial equilibrium model for the electricity and combined heat and power sectors, Balmorel is presented. In accordance with the classification undertaken earlier in this chapter, Balmorel can be classified as a supply-side energy system model that covers the electricity and coupled energy sectors in part. The main strength of this model is its high modularity, enabling analyses at various geographical levels (regional, national, international) and different additional sectors like electric mobility, individual heating and hydrogen production. The model was one of the first open-source energy system models in Europe. It was originally developed in the course of the Balmorel Open Source Project, as documented by Ravn et al. (2001), Ravn (2001) and Grohnheit and Larsen (2001). It has now been further developed by a wide range of institutions within Europe and worldwide, e.g., universities and research institutions, regulatory and energy authorities, transmission system operators, consultants and energy companies. Figure 3.2 shows a schematic overview of the Balmorel model. Balmorel is implemented in GAMS and uses linear and mixed-integer programming to maximise social welfare subject to technical, physical and regulatory constraints representing the overall energy system’s cost (fuel, transmission, fixed and variable O&M costs, taxes and subsidies, minus consumers’ utility) while satisfying the electricity and heat demand

Energy system modelling of renewable energy  47

Source:   Wiese et al. (2018), CC BY 4.0

Figure 3.2  Balmorel schematic overview (Wiese et  al. 2018). The objective function (Equation 3.1) minimises discounted system fom costs split into investment costs (cinv y ), fixed operation and maintenance costs (c y ) and varivom able operation and maintenance costs (cy ) weighted by the discount factor (DFy ) (GeaBermúdez et al. 2021):

min

fom vom cinv y , cy , cy

åDF * ( c y

inv y

)

+ cyfom + cyvom (3.1)

y

The model is restricted by further equations on electricity and district heat balance, capacity and energy constraints, the production of dispatchable and non-dispatchable units, operational constraints, storage operation, transmission constraints, emission caps and several more. The deterministic energy system model combines top-down and bottom-up elements and optimises energy dispatch and (optionally) investments in generation units, storage, transmission or other technology components. Table 3.1 gives an overview of the required exogenous input parameters and expected output parameters in the dispatch and investment optimisation in Balmorel. Annual demand (district heating and electricity) and corresponding profiles are exogenous inputs. The temporal resolution can be chosen according to the research focus but is in most cases hourly resolution. The outputs of the model are electricity and heat generation, fuel consumption, electricity exports and imports, emissions, investments in plants and transmission lines, prices of traded energy and total costs. The model runs can be conducted either in perfect foresight or rolling horizon mode. Electric and thermal storage can be modelled as well.

48  Handbook on the economics of renewable energy

Table 3.1  Input and output parameters for dispatch and investment optimisation in Balmorel Input

Output

Dispatch optimisation • Electricity and district heat demand and hourly profiles • Fuel prices • Generating capacities and their characteristics • Resource characteristics for wind, hydro and solar resources • Transmission capacities and transmission and distribution losses

• Electricity and district heat generation per generation unit • Electricity distribution and transmission • System cost • System emissions

Investment optimisation • Investment cost for different technology types • Investment cost for transmission capacity • Interest rate • Economic lifetime of technologies

• Endogenously installed generation capacity per technology type • Endogenously installed transmission capacity between regions

Table 3.2  Balmorel model characteristics Balmorel model characteristics System aggregation

Flexible at three levels (spatial levels of countries, regions and areas)

Optimisation type

Linear programming

Optimisation focus

Minimising annualised costs of the energy system

Optimisation object

Dispatch and investment

Output

Energy production by unit, fuel consumption, emissions, electricity import/export, investments in plants and transmission, as well as electricity price

Model run-time

Depending on the size of the problem, varying from minutes to days

Access

Complex interface, open source (demands GAMS license and linear programming software), direct access to code and data

Source:   adapted from Münster (2019)

Spatial modularity allows for the modelling of different regions that can represent, e.g. different pricing zones within a country or transmission constraints, and respective interregional or international trade within the system boundaries in the model. Policy restrictions in terms of emission targets or decarbonisation goals are also implemented. In the Balmorel model, an electricity price is calculated for each region and each time segment of the modelled period. This price represents the electricity producers’ marginal cost of generation (including fuel costs, fuel and emission taxes, operation and maintenance costs and investment costs). Table 3.2 shows the model characteristics of Balmorel. The openly available code and data are accessible on the internet under an ISC license (Balmorel 2022). The collaboration of various groups in the development of Balmorel as well

Energy system modelling of renewable energy  49

as the wide range of applications increase the robustness of the model. The analyses that have been conducted with Balmorel cover a wide geographical scope and range from regional case studies focusing on the electricity sector (for example Barragán-Beaud et al. 2018; Fedato et al. 2019; Schöniger et al. 2021) to the coverage of several countries and sectors, including the electricity, heat and transport sector (e.g., Gea-Bermúdez et al. 2021; Jensen and Skovsgaard 2017; Nagel et al. 2022). Enertile: An Energy System Optimisation Model Offering a High Granularity of Renewable Energy Potentials As mentioned above, Enertile is an energy system optimisation model representing the electricity, grid-connected heat and hydrogen sectors (Fraunhofer ISI 2022) in high temporal, spatial and techno-economic resolution. Enertile optimises in an integrated manner the dispatch of and investment in the supply technology options for the generation, conversion, transmission and distribution of the described sectors. The model has been used for a wide range of mostly long-term studies, often dealing with the integration of high shares of renewable energies into the energy system. Sector coupling like residential heat supply by heat pumps and electric mobility can be implemented as well. A strength of Enertile is its integrated method for the calculation of renewable energy potentials and generation profiles in very high granularity, which considers detailed regional weather input data and potential limitations like distance regulations and protected areas (Fraunhofer ISI 2022). The model includes renewable and conventional generation units, storage and grid capacities for these sectors. Electricity, heat and hydrogen demands are given exogenously. Figure 3.3 shows the structure of the Enertile model. Supply and demand are matched on an hourly basis, including the representation of storage and transmission grids. The output data in hourly and annual resolution cover the generation, fuel consumption, CO2 emissions, costs (capital, fuel, operation and maintenance) and endogenous shadow prices for electricity as well as CO2 (Fraunhofer ISI 2022). Table 3.3 shows the main model characteristics of Enertile (v5.1). Green-X: A Tool for Techno-Economic RE Policy Evaluation The model Green-X was originally developed by the Energy Economics Group at TU Wien within the homonymous EU research project in the period 2002 to 2004. Initially focussed on the electricity sector, this modelling tool and its database on renewable energy potentials and costs have been extended to incorporate renewable energy technologies within all energy sectors. Green-X covers the European Union and its member states, the contracting parties of the Energy Community (the Western Balkans, Ukraine, Moldova) and selected other EU neighbours (Turkey, North African countries). It allows the investigation of the future deployment of RE as well as the accompanying costs (including capital expenditures, additional generation cost of RE compared to conventional options and consumer expenditures due to applied supporting policies) and benefits (for instance, avoidance of fossil fuels and corresponding carbon emission savings). Results are calculated at both a country and a technology level on a yearly basis. The time horizon allows for in-depth assessments up to 2050. The Green-X model develops nationally specific dynamic cost-resource curves for all key RE technologies,

50  Handbook on the economics of renewable energy

Source:   Fraunhofer ISI (2022)

Figure 3.3   Enertile structure including for renewable electricity, biogas, biomass, biowaste, on- and offshore wind, hydropower large- and small-scale, solar thermal electricity, photovoltaic, tidal stream and wave power and geothermal electricity; for renewable heat, biomass, sub-divided into log wood, wood chips, pellets, grid-connected heat, geothermal grid-connected heat, heat pumps and solar thermal heat; and, for renewable transport fuels, first-generation biofuels (biodiesel and bioethanol) and second generation biofuels (lignocellulosic bioethanol, biomass to liquid), as well as the impact of biofuel imports. Besides the formal description of RE potentials and costs, Green-X provides a detailed representation of dynamic aspects such as technological learning and technology diffusion. Through its in-depth energy policy representation, the Green-X model allows an assessment of the impact of applying (combinations of) different energy policy instruments (for instance, quota obligations based on tradable green certificates/guarantees of origin, (auctioned) feed-in

Energy system modelling of renewable energy  51

Table 3.3  Enertile model characteristics for model version 5.1 Scope Resolution of results for renewable technologies

Germany, Europe, MENA and flexible regions Country level, grid 11 kW) and slightly

A mixed-integer linear programming approach  351 DN60 150 m

DN60 50 m

DN60 100 m

Figure 15.3  Four-node DHS with thermal storage, a large-scale heat pump and a lowgrade heat source

Figure 15.4  Heat flow for the four-node system at the production unit over time higher opex. The figure shows higher deviations during rapid changes in the heat flow at the production unit. The deviation is caused by higher heat losses (>5%) in the nonlinear model due to a higher temperature Taout ,t (see Equation 15.6). The standard deviation is 7.5 kW (= 1.21%). The estimated velocity vaest was set to 0.16  m  s–1 based on the nonlinear model results. For higher velocities, the heat losses decrease, causing lower nominal heat flows for the supply system. The results led to the discussion focused on dealing with the error due to falsely calculated heat losses and the impact of neglecting hydraulics. The estimated velocity and the heat transfer are the influential parameters for the heat losses (see Equation 15.10). If a DHS is to be optimized, we suggest fitting those two parameters to existing data depicting reality as close as possible. Another significant parameter is the diameter which causes problems in the

352  Handbook on the economics of renewable energy

solving stage due to the violation of Equation 15.7. Therefore, profound knowledge of the grid and its consumers is essential for optimizing the supply system. The results suggest that the error is solely caused by the different heat losses causing different nominal heat flows at the supply system. This suggestion leads to the assumption that hydraulics do not affect the design of the supply system. Two arguments support the assumption: Equation 15.7 is a hydraulic boundary for the heat flow at 250 Pa m–1 ensuring a maximum pressure loss in the pipes. The second argument is the low investment for pumps compared to energy converters. The Grundfos Hydro MPC-E 2 CRE 95-2-2 costs €71 723.71 and, at the optimal operation point, generates a flow rate of 164.3  m3  h–1 with a power of 12.08 kW (Grundfos, 2022). Regarding power per cost, the pump is more expensive than a large-scale heat pump (heat source: lake/sea) (Steinbach et al., 2020). However, in absolute costs, a 1 MW large-scale heat pump would be 6.7 times more expensive than the pump. In addition, the operating power of 12.08  kW is insignificant compared to 1  MW. The small differences between the linear and nonlinear model results further verify the linear model’s accuracy. If a high level of technical detail is required, we recommend a detailed simulation applying tools like TRNSYS (Transsolar Software Engineering, 2021) or PandaPipes (Cronbach et al., 2022) after the optimization. However, the achieved accuracy should usually be sufficient at the planning stage. Introduction of the Case Study After validating the linear model, we applied it to a district heating grid in eastern Germany using 2021 as the base year with a time horizon of ten years. For the case study, we considered one-third of the grid, measuring a length of 4 km and an annual demand of 3.27 GWh. The grid’s maximum capacity based on a 20 K temperature difference between the current timestep and the following one is 2 GWh. The supply system connects to the upper pipe in Figure 15.5, supplying 26 transfer stations through the grid. The figure illustrates the grid, the buildings and the supply system’s integration point. Additionally, the consumer’s demand summation increases in the colder months and drops significantly in the summer due to the lack of space heating. The linear optimization algorithm is applied to the grid to design the supply system at the given integration point. The design is primarily based on energy converters and carriers’ economic data. The economic data for CHP biogas/gas, an industrial heat pump and a biomass plant can be found in Peters et al. (2022) for specific capex and in Verein Deutscher Ingenieure (2012) for the remaining data. Dynamic Storage Behaviour of the Grid For the dynamic storage behaviour, we optimized with a variable electricity price on the spot market for every hour in 2021. The question is whether the grid’s storage behaviour is a significant variable to be optimized and affects the design optimization of the supply system. For this research, an industrial heat pump with a constant heat source was selected, providing a temperature level of 60°C. The results in Figure 15.6 show the heat flow at the supply’s integration point with a maximum capacity of 1350 kW in the upper plot with and without the grid’s storage capacity. The

A mixed-integer linear programming approach  353

Figure 15.5  One-third of a DHS in eastern Germany was used for the case study with the summed demand of the consumers in the grid over the year 2021 heat pump control seems quite similar, and the standard deviation over the year is 65 kW. Furthermore, the plot illustrates the charging and discharging of the grid and compares this behaviour to the current electricity price. In this plot, an explicit dependency between charging and discharging the grid and the electricity price cannot be identified. In addition, the heat generation costs only differ by 0.3 ct (kW h) –1. The discussion focuses on why the grid’s storage capacity has almost no influence on the costs. Looking at Figures 15.5 and 15.6, we can identify a significant gap between generated and consumed heat. This gap is caused by the grid’s heat losses, about 20% of the total heat generation (neglecting the transfer station’s efficiency). These high heat losses are caused by high temperatures with an average temperature of 83°C over the year and the heat transfer coefficient. Due to the high losses, the optimizer minimizes them by lowering the temperature in the grid if possible. Therefore, the optimizer controls the grid based on the temperature required rather than the energy carrier prices. This results in the finding that lowering the temperature in such high-temperature grids should be the top priority in the transformation process. Measures to lower temperatures include improving pipe insulation, refurbishing buildings and decentralizing units to decouple or support high-temperature consumers. Additionally, the grid’s storage capacity is still essential when modelling sector-coupled systems because power-to-heat technologies can help to balance the electric grid in critical times. The depiction of dependencies on the electric grid requires additional boundaries in the model. Additionally, it should be noted that with decreasing heat losses, the need for controlling power-to-heat technologies increases depending on the electric market.

354  Handbook on the economics of renewable energy

Figure 15.6  The upper plot shows the thermal power at the integration point of the supply system for the model with the grid’s storage capacity and without the grid’s storage capacity; the lower plot depicts the charging and discharging of the grid and the electricity price over time Economic Assessment of Renewable and Non-Renewable Technologies in District Heating Systems To further examine the potential of large-scale heat pumps, we compared them with a CHP plant based on gas and biogas and a heating plant fired by biomass. The assessment is based on the year 2021; however, a gas scenario in the year 2022 was examined as well. The opex mainly depend on the energy carrier price, and for this research, we assumed constant energy carrier prices: ● ● ● ● ● ● ●

Gas 2021: 6.7 ct (kW h) –1 (Statistisches Bundesamt, 2021) Gas 2022: 12.21 ct (kW h) –1 (Statistisches Bundesamt, 2021) Biogas 2021: 8.2 ct (kW h) –1 (IWB, 2021) Biomass 2021: 3.6 ct (kW h) –1 (CARMEN EV, 2021; Müller, 2022) Electricity 2021 buy: 19.2 ct (kW h) –1 (eex, 2022) Electricity 2021 sell: 9.7 ct (kW h) –1 (eex, 2022) Electricity 2022 sell: 11.6 ct (kW h) –1 (eex, 2022)

The electricity price in 2021 for selling energy was applied to the CHP plants that can produce heat and electric power. For gas, a subsidy of 5.4 ct (kW h) –1 is applied based on KWKG (the German CHP) law and grid relief charges. Based on EEG law and grid relief charges, a fixed selling price of 13.54 ct (kW h) –1 is applied for biogas. Furthermore, investment subsidies and the EEG subsidy for electricity prices are not included. Table 15.2 includes all relevant techno-economic data used in the optimization. The technology-specific data were extracted from Peters et  al. (2022) and Verein Deutschter

355

2.5

Large-scale heat pump

Source:   Verein Deutscher Ingenieure (2012)

Heat grid

6

Heat plant biomass

8

CHP gas 2022

6

8

CHP gas 2021

CHP biogas 2021

Opex fixed [%]

Technology

0.12

Opex variable [ct (kW h) –1]

19.2

3.6

8.2

12.21

6.7

Energy carrier [ct (kW h) –1]

Table 15.2  Techno-economic parameters for the technologies

761

902

790

720

720

Capex [€ (kW) –1]

13.54

17

15.1

Selling energy [ct (kW h) –1]

20

15

15

15

15

Lifetime [a]

~ 0.8(th)

~ 5.9(th)

0.8(th)

0.391(th) 0.379(el)

0.408(th) 0.379(el)

0.408(th) 0.379(el)

Efficiency

356  Handbook on the economics of renewable energy

Ingenieure (2012). The fixed opex are expressed as a percentage of the capex. For the comparison, a base scenario, scenarios for the electricity price and scenarios for the heat pump’s heat source were performed. Due to EEG, the electricity price changes do not affect the selling price for biogas-based electricity because the law guarantees a fixed price. The left plot in Figure 15.7 illustrates the results of the base scenario. The plot shows the five technologies’ cost development over the optimized year. Furthermore, the heat generation costs were calculated. All cost curves’ gradients decrease towards the summer months due to lower demand. The large-scale heat pump has the cheapest yearly capex-related annuity due to its low specific capex and its longer lifetime. The CHP plants follow with a specific capex of 720 € (kW) –1 (gas-fired) and 790 € (kW) –1 (biogas-fired) with a lifetime of 15 years. The biomass heat plant has the highest capex due to 902 € (kW) –1. The costs of the CHP gas plant from 2022 increased rapidly due to a much higher energy carrier price. This price increase leads to the highest heat generation costs of 33.7 ct (kW h) –1. The large-scale heat pump has the second lowest heat generation costs with 11 ct (kW h) –1, but the highest opex. The low heat generation costs result from a high COP (~5.9) caused by a high heat source temperature (60°C). The CHP plant fired by gas in 2021 has slightly lower heat generation costs of 10.9 ct (kW h) –1. The biomass heat plant costs 15.7 ct (kW h) –1, and the biogas-fired CHP plant costs 22.1 ct (kW h) –1. The plot on the right-hand side of Figure 15.7 displays the cost development for the heat source scenario. The different heat source temperatures result in different COPs causing lower costs for increased efficiency and higher costs for decreased efficiency. The temperature was decreased and increased by 10  K compared to the base scenario. The 50°C and

Figure 15.7  Base scenario (left), heat source scenario (right) – absolute cost development in k€ over one year for CHP plants based on gas 2021, gas 2022, biogas and a large-scale heat pump (HP in the figure) and biomass on the left-hand side; the right-hand side is the heat source scenario; in the legend, the heat generation costs for each technology are included as well as a CHP plant with gas 2021 as a comparison

A mixed-integer linear programming approach  357

70°C heat sources develop similarly to the base heat source varying by around 1.5 ct (k Wh) –1 in heat generation costs. The 5°C heat source was chosen as an extreme case resulting in 26.5 ct (k Wh) –1 with an average COP of 1.78. The gas-fired CHP plant starts with a higher initial value than the heat pumps but generates heat for a lower price, except for the heat pump with a 70°C heat source. The electricity price scenario is illustrated in Figure 15.8, with the electricity price decreasing and increasing by 10%. The electricity price is increased by 10% on the left-hand side, and on the right-hand side, the price is decreased. The CHP plants’ costs decrease on the left-hand side due to a higher selling price for electricity, except for biogas-based electricity because of a fixed price. The heat pump is slightly more expensive, with a heat generation price of 11.4 ct (kW h) –1. The same developments can be seen on the right-hand side, only vice versa. The scenario discussion focuses on the validity of results and economical renewable DHS in the future. For the demand input no concurrency factor was applied resulting in a more robust system with higher costs. The heat generation costs of the gas-fired CHP plant in 2022 are incredibly high. This is caused by the change in the ratio between the gas and the electricity price. The gas price has doubled while the electricity price has only increased by 20%. In addition, the result of the costs is a price that would be paid for the next ten years. In a static cost calculation without the annuity, the gas-fired CHP plant in 2022 would cost about 28  ct  (kW  h) –1. The biomass heat plant generation costs are at 15.7  ct  (kW  h) –1 and 6.8 ct (kW h) –1, which are higher compared to Schuster et al. (2017). However, biomass prices have increased significantly, and the study was published five years ago. Compared to a gasfired CHP plant in 2022, a biomass heat plant is still competitive.

Figure 15.8  Electricity price scenario – absolute cost development in k€ over time for CHP plants based on gas 2021, gas 2022, biogas and a large-scale heat pump (HP in the figure), and biomass with a 10% higher electricity price on the left-hand side; the right-hand side is analogous to the left-hand side with a 10% lower electricity price

358  Handbook on the economics of renewable energy

The large-scale heat pump has the second-best economic performance compared to the other technologies and only is slightly beaten by 0.1 ct (kW h) –1 compared to the gas-fired CHP plant in 2021. The low heat generation costs are mainly due to a very high COP (5.9 on average) caused by a high-temperature heat source. Arpagaus et al. (2018) stated that the Ochsner IWWDS ER3c4 has a COP of 5.9, lifting the refrigerant’s temperature from 65 to 90°C. In this study, the average COP is calculated based on a temperature lift of 23 K from 60 to 83°C. In addition, the Carnot efficiency for this specific case equals 15.5. Therefore, the calculated COP based on Jesper et al. (2021) seems reasonable. However, heat pumps are limited in the temperature lift and the sink temperature. Arpagaus et al. (2018) reviewed different large-scale heat pumps and stated a technical limitation for commercial heat pumps of a 90 K temperature lift. This technical limitation might cause problems if low-grade heat sources supply heat pumps in high-temperature DHS, and a cascade between several heat pumps can solve this problem. Shallow geothermal energy would be an example of a low-grade heat source supplying temperatures around 5 to 10°C in the winter. Therefore, implementing temperature lowering measures is crucial for transforming DHS. For high-temperature DHS, heat sources with a high temperature level are best suited in combination with a large-scale heat pump. Typically those heat sources come from industrial waste heat. Papapetrou and Kosmadakis (2022) identified the paper and pulp industry as a potential waste heat provider supplying temperatures of 70–80°C. Another option is the food and beverages industry, where the waste heat temperatures can vary from 60°C to 500°C but are typically below 200°C (Papapetrou and Kosmadakis, 2022). In general, the waste heat potential in Europe is 304.13  TW  h  a–1 (Papapetrou and Kosmadakis, 2022) which is two-thirds of Germany’s space heating demand in 2020 (Statista, 2022). Most industries can supply waste heat temperatures above 40°C and are attractive heat sources for large-scale heat pumps supplying heat into a DHS. Of course, there needs to be an incentive for the companies, and depending on the temperature level and the availability, the company could set the price for selling its waste heat. In the base scenario with a temperature level of 60°C, the grid operator could still be competitive if it bought the waste heat for 1–2 ct (kW h)–1 compared to biomass or CHP plants – gas prices below 7 ct (kW h) –1 would favour a CHP plant. Additional information can be extracted by viewing the emissions compared to the heat generation costs in Figure 15.9. The emission calculation is based on the total generated energy of 5.16 GW h. The heat generation costs are taken from the base scenario, and the emission factors were based on the German market (German electricity mix). The biomass heat plant and the large-scale heat pump achieved the best overall performance in terms of costs and CO2 emissions. The biomass plant has lower emissions but higher costs, while the heat pump has lower costs but higher emissions. The large-scale heat pump produces 418 t CO2 due to the German electricity mix consisting of 55% non-renewable generation (Holm, 2021). The CHP plant based on gas in 2022 generates heat with the highest costs and CO2 emissions. The discussion tackles the question of whether combining the heat pump with industrial waste heat and biomass heat plants is the solution to the transformation of DHS or whether CHP plants based on biogas can also play a role in a renewable energy system. The emissions support the argument for a heat pump and biomass-based system; however, it should be noted that CHP plants can also produce electricity and support the grid’s stability during heat peak load times. In Germany, this stability increase is benefitted by the reduction of emissions. On top of that, combining CHP plants and heat pumps might be essential in the future and even lead to an economical competitive system due to rising electricity prices and the ability to produce electricity and directly convert it into heat. The flexibility needs to increase if more large-scale heat pumps are connected to the

A mixed-integer linear programming approach  359

Source:   Peters et al., 2022

Figure 15.9   CO2 emissions and heat generation costs for CHP plants based on gas in 2021/2022, and biogas and heat plant based on biomass and a large-scale heat pump electric grid. As explained, storing thermal energy in the grid or storage is one key factor; another one is the combination of heat pumps with different technologies that can produce electricity. Besides the discussion of combining biogas, biomass and heat pumps, the results clearly indicate that gas utilization in DHS is neither an economic nor sustainable solution.

LIMITATION OF THE RESULTS The optimization algorithm introduced can design complex supply systems in district heating with multivalent operation. However, for the economic assessment, the optimizer was limited to choosing from one technology at a time. This restriction limits the possible outcomes in order to make a direct comparison between the chosen technologies. Usually, the system would be supplied by a combination of technologies. Additionally, the waste heat source payment was not included due to the lack of information on the regular pricing of waste heat sources. Therefore, possible prices for waste heat were suggested in the discussion, comparing the waste heat prices to established technologies such as gas-fired CHP plants.

CONCLUSION The main findings of this work can be divided into two categories: modelling DHS and economic assessment of four technologies. In summary, modelling the hydraulics of the grid has almost no effect on the design, with an economic deviation of less than 1%. The same

360  Handbook on the economics of renewable energy

result was found in modelling the dynamic storage capacities of the grid. The optimizer could have used a fluctuating electricity price to save costs by injecting the heat into the grid when the energy costs decrease. However, the heat losses increase with an increasing temperature, resulting in higher costs than the savings due to the grid’s storage capacity. Therefore, optimizing the temperature in high-temperature DHS should be critical in the transformation phase. The variation of the energy carrier price for designing DHS has a lower impact compared to temperature lowering measures. Nevertheless, the grid’s storage capacity is essential in sector-coupled systems to balance the electric grid. In addition, seasonal storage can play a significant role in the transformation of DHS due to the capability of storing large amounts of energy when the costs are low and the outside temperature is high. Besides the influence of heat losses on economics, the supply system significantly impacts heat generation costs. The research has shown that large-scale heat pumps combined with waste heat are competitive with heat generation costs of 12 ct (kW h) –1 (including the pricing for waste heat) compared to CHP plants with generation costs of 10.9 ct (kW h) –1 (the gas price of 2021) and 33.7 ct (kW h) –1 (the gas price of 2022). Biomass will also be an economical option in renewable DHS with an estimated price of 15.7 ct (kW h) –1, especially because biomass has the lowest emissions. In 2022, gas-fired CHP plants are no longer competitive due to the rising gas prices. Biogas-fired CHP plants were more expensive than the other two renewable technologies, at 22.1 ct (kW h) –1. However, it might be an option if no heat source at a high temperature level is available and a renewable share of the heat is obligatory. However, shifting to renewable energy systems will require multivalent and sector-coupled systems. The findings implicate that future renewable DHS should be operated at low temperatures and utilize waste energy with heat pumps to increase their economic benefits. Furthermore, those systems should be operated multivalently and use storage facilities to react on the electrical grid. Low-temperature DHS can already compete with fossil-based alternatives. In particular, large-scale heat pumps need to be part of a renewable DHS combining it with solar thermal energy and other renewables discussed in “Techno-Economic Background of District Heating Systems”. Therefore, further research is needed in designing sector-coupled energy systems on a local scale and developing temperature-lowering measures for DHS. The utilization of different storage technologies, such as seasonal storage and the coupling to the electric grid, needs to be further examined. Additionally, combining different heat sources and generators needs further investigation to support planners and operators in transforming their DHS. In the end, a renewable DHS will have significant economic benefits if the grid’s temperature is lowered.

REFERENCES Arpagaus, C., Bless, F., Uhlmann, M., Schiffmann, J., and Bertsch, S. S. (2018). ‘High temperature heat pumps: Market overview, state of the art, research status, refrigerants, and application potentials’. Energy, vol. 152, pp. 985–1010. https://doi​.org​/10​.1016​/j​.energy​.2018​.03​.166. Best, R. E., Rezazadeh Kalehbasti, P., and Lepech, M. D. (2020). ‘A novel approach to district heating and cooling network design based on life cycle cost optimisation’. Energy, vol. 194, p. 116837. Bestuzheva, K., Besançon, M., Chen, W.-K., Chmiela, A., Donkiewicz, T., van Doornmalen, J., Eifler, L., Gaul, O., Gamrath, G., Gleixner, A., Gottwald, L., Graczyk, C., Halbig, K., Hoen, A., Hojny, C., van der Hulst, R., Koch, T., Lübbecke, M., Maher, S. J., Matter, F., Mühmer, E., Müller, B., Pfetsch, M. E., Rehfeldt, D., Schlein, S., Schlösser, F., Serrano, F., Shinano, Y., Sofranac, B., Turner, M., Vigerske, S., Wegscheider, F., Wellner, P., Weninger, D., and Witzig, J. (2021). ‘The SCIP

A mixed-integer linear programming approach  361

Optimization Suite 8.0 [Online]’. Zuse Institute Berlin. Available at: https://optimization​-online​.org​/ 2021​/12​/8728/ (Accessed 31 August 2022). Bornand, B., Girardin, L., Belfiore, F., Robineau, J.-L., Bottallo, S., and Maréchal, F. (2020). ‘Investment planning methodology for complex urban energy systems applied to a hospital site’. Frontiers in Energy Research, vol. 8, p. 537973. C.A.R.M.E.N. E.V. (2021). ‘Marktpreise Hackschnitzel [Online]’. Available at: https://www​.carmen​ev​.de​/service​/marktueberblick ​/marktpreise​-energieholz​/marktpreise​-hackschnitzel/  (Accessed  10 August 2022). Corscadden, J., Möhring, P., and Krasatsenka, A. (2021). ‘Transformation of existing urban district heating and cooling systems from fossil to renewable energy sources, Euroheat & Power and Hamburg Institut [Online]’. Available at: https://www​.res​-dhc​.com​/wp​-content​/uploads​/2021​/05​/ RES​-DHC​_WP2​_Task​_2​.1​_D2​.1​_ Survey​_ EU​-Level​_ FINAL ​_UPDATED​_202104​.pdf (Accessed 9 May 2022). Cronbach, D., Lohmeier, D., Drauz, S. R., Kisse, J., and Kneiske, T. (2022). ‘PandaPipes [Online], Fraunhofer Institute for Energy Economics and Energy System Technology IEE’. Available at: https://www​.pandapipes​.org/. Delubac, R., Serra, S., Sochard, S., and Reneaume, J.-M. (2020). ‘A multi-period tool to optimise solar thermal integration in district heating networks’. In: ECOS 2020 – Proceedings of the 33rd International Conference on Efficiency, Cost, Optimisation, Simulation and Environmental Impact of Energy Systems [Online]. Available at: https://www​.scopus​.com​/inward​/record​.uri​?eid​=2​-s2​.0​85095788908​&partnerID​= 40​&md5​= 479​c416​b2b3​be75​5a91​bc73​79951217f. eex. (2022). ‘Strommarkt [Online]’. Available at: https://www​.eex​.com​/de​/marktdaten​/strom (Accessed 17 August 2022). European Commission. (2021a). ‘Fernwärmenetze - Kapazitäten, Erzeugung und Verluste nach Netztyp und Anlagentyp [Online]’. Available at: https://appsso​.eurostat​.ec​.europa​.eu​/nui​/show​.do​?dataset​ =nrg​_dhdc​_cpl​&lang​=de (Accessed 1 June 2022). European Commission. (2021b). ‘Gross derived heat generation by fuel, EU, 2000-2019 (GWh) [Online]’. Available at: https://ec​.europa​.eu​/eurostat​/statistics​-explained​/index​.php​?title​=File​:Gross​ _derived​_heat​_generation​_by​_fuel,​_ EU,​_2000​-2019_(GWh)​_T4​.png (Accessed 1 June 2022). Federal Ministry of Economics and Energy. (2019). ‘Energiedaten: Gesamtausgabe [Online]’. Available at:  https://www​.bmwi​ .de ​ / Redaktion ​/ DE ​ / Downloads ​ / Energiedaten ​/energiedaten ​ -gesamt​ -pdf​grafiken​.pdf?_​_blob​=publicationFile​&v​=34 (Accessed 8 March 2021). Grundfos. (2022). ‘Hydro MPC-E 2 CRE 95-2-2 Modellnummer 99441150 [Online]’. Available at: https://product​-selection​.grundfos​.com ​/de​/products​/ hydro​-mpc​/ hydro​-mpc​-e​/ hydro​-mpc​-e​-2​-cre​-95​ -2​-2​-99441150​?pumpsystemid​=1621194926​&tab​=variant​-sizing​-results (Accessed 28 July 2022). Holm, L. M. (2021). ‘Strommix & Stromerzeugung [Online], Strom-Report’. Available at: https://strom​ -report​.de​/strom/ (Accessed 18 August 2022). IWB. (2021). ‘Tarife für die Lieferung von Biogas-Erdgas [Online]’. Available at: https://www​.iwb​.ch​ /dam​/jcr​:97697938​-3382​- 4438​-8044​-14fdf3c75f48​/ IWB​%20Biogas​-Erdgas​%20Tarifblatt​%20mit​ %20MWST​%200120​%20v2​.pdf (Accessed 10 August 2022). Jesper, M., Schlosser, F., Pag, F., Walmsley, T. G., Schmitt, B., and Vajen, K. (2021). ‘Large-scale heat pumps: Uptake and performance modelling of market-available devices’. Renewable and Sustainable Energy Reviews, vol. 137, p. 110646. Konstantin, P. (2017). Praxisbuch Energiewirtschaft: Energieumwandlung, -transport und -beschaffung, Übertragungsnetzausbau und Kernenergieausstieg [Online]. 4th edn. Berlin, Germany: Springer Berlin Heidelberg. Available at: https://ebookcentral​.proquest​.com​/ lib​/gbv​/detail​.action​?docID​= 4813396 (Accessed 23 November 2020). Krug, R., Mehrmann, V., and Schmidt, M. (2020). ‘Nonlinear optimisation of district heating networks’. Optimization and Engineering, vol. 19, no. 1, p. 1. Li, H., and Svendsen, S. (2013). ‘District heating network design and configuration optimisation with genetic algorithm’. Journal of Sustainable Development of Energy, Water and Environment Systems, vol. 1, no. 4, pp. 291–303 [Online]. DOI: 10.13044/j.sdewes.2013.01.0022. Lund, H., Østergaard, P. A., Nielsen, T. B., Werner, S., Thorsen, J. E., Gudmundsson, O., Arabkoohsar, A., and Mathiesen, B. V. (2021). ‘Perspectives on fourth and fifth generation district heating’. Energy, vol. 227, p. 120520. DOI: 10.1016/j.energy.2021.120520.

362  Handbook on the economics of renewable energy

Mavromatidis, G., and Petkov, I. (2021). ‘MANGO: A novel optimisation model for the long-term, multi-stage planning of decentralised multi-energy systems’. Applied Energy, vol. 288, p. 116585. Müller, A. (2022). ‘Pelletpreise – Deutschlands Heimvorteil sorgt für Stabilität [Online], DAA GmbH’. Available at: https://www​.heizungsfinder​.de​/pelletheizung​/pellets​/preise​-kosten#:~​:text​ =In​%20der ​%20ersten​%20Januarh​%C3​%A4lfte​%202022​%20lag​%20der ​%20Pelletpreis​,Tonne.​ %20Preisentwicklung​%20bei​%20Pellets​%20von​%202012​%20bis​%202022 (Accessed 17 August 2022). Nussbaumer, T., Thalmann, S., Ardens, A. J., and Ködel, J. (2018). ‘Planungshandbuch Fernwärme, Consortium QM District Heating [Online]’. Available at: www​ .qmfernwaerme​ .ch (Accessed 23 November 2020). Paardekooper, S., Lund, R. S., Mathiesen, B. V., Chang, M., Petersen, U. R., Grundahl, L., David, A., Dahlbæk, J., Kapetanakis, I. A., Lund, H., Bertelsen, N., Hansen, K., Drysdale, D. W., and Persson, U. (2018). ‘Heat Roadmap Europe 4: Quantifying the Impact of Low-Carbon Heating and Cooling Roadmaps, Aalborg University [Online]’. Available at: https://vbn.aau.dk/ws/portalfiles/ portal/288075507/Heat_Roadmap_Europe_4_Quantifying_the_Impact_of_Low_Carbon_ Heating_and_Cooling_Roadmaps.pdf (Accessed 9 May 2022). Papapetrou, M., and Kosmadakis, G. (2022). ‘Resource, Environmental, and Economic Aspects of SGHE’. In Salinity Gradient Heat Engines. Elsevier, pp. 319–353. Peters, M., Steidle, T., Hebisch, H., Skok, J., Berg, A., Graef, D., and Anders, F. (2022). ‘Einführung in den Technikkatalog zur kommunalen Wärmeplanung in Baden-Württemberg [Online], Stuttgart, Ministry of the Enviornment, Climate Protection and the Energy Sector (Ministerium für Umwelt, Klima und Energiewirtschaft) - Baden-Württemberg’. Available at: https://um​.baden​-wuerttemberg​ .de​/fileadmin​/redaktion​/m​-um​/intern​/ Dateien​/ Dokumente​/2​_ Presse​_und​_ Service​/ Publikationen​/ Energie​/ Kommunale​-Waermeplanung​-Einfuehrung​-in​- den​-Technikkatalog​-Version​-1​-barrierefrei​ .pdf (Accessed 21 July 2022). Schuster, M., Maier, P., Bargmann, E., and Bader, H. (2017). ‘Investitionen in Biomasse-Wärme für Dritte: Mit Waldhackgut Einkommen sichern [Online], Wien, Bundesministerium für Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft’. Available at: https://www​.klimaaktiv​.at​/dam​/jcr​ :2bcbcd60​- 6a5d​- 4cd7​-96da​-a09835eff8b6​/ Fachinformation​_ Biomasse​-W​%C3​%A4rme​_ FINAL​ .pdf (Accessed 18 August 2022). Sporleder, M., Burkhardt, M., Kohne, T., Moog, D., and Weigold, M. (2020). ‘Optimum Design and Control of Heat Pumps for Integration into Thermohydraulic Networks’. Sustainability, vol. 12, no. 22, p. 9421. DOI: 10.3390/su12229421 (Accessed 23 November 2020). Stange, P., and Matthees, A. (2020). ‘Projekt “DELFIN”: FreePlan Version 1.1 [Online], TU Dresden’. Available at: https://tu​-dresden​.de​/ing​/maschinenwesen ​/iet ​/gewv​/forschung​/forschungsprojekte​/ delfin (Accessed 25 July 2022). Statista. (2022). ‘Energieverbrauch für Wärmezwecke in Deutschland nach Sektoren im Jahr 2020 [Online]’. Available at: https://de​.statista​.com​/statistik​/daten​/studie​/614202​/umfrage​/ waermeverbrauch ​ -in ​ - deutschland ​ - nach ​ - sektoren/#:~​:text​ = Die​ %20Statistik ​ %20zeigt ​ %20den​ %20Energieverbrauch ​ , Petajoule ​ % 20Energie ​ % 20f ​ % C3​ % BCr ​ % 20Raumw ​ % C3​ % A4r me​ %20verbraucht (Accessed 19 August 2022). Statistisches Bundesamt. (2021). ‘Erdgas- und Stromdurchschnittspreise [Online]’. Available at: https:// www​.destatis​.de​/ DE​/ Themen ​/ Wirtschaft ​/ Preise​/ Erdgas​-Strom​-DurchschnittsPreise/​_ inhalt​.html​ #sprg475782 (Accessed 17 August 2022). Steinbach, J., Popovski, E., Henrich, J., Christ, C., Ortner, S., Pehnt, M., Blömer, S., Auberger, A., Fritz, M., Billerbeck, A., Langreder, N., Thamling, N., Sahnoun, M., and Rau, D. (2020). ‘Umfassende Bewertung des Potenzials für eine effiziente Wärme- und Kältenutzung für Deutschland, IREES, ifeu, Fraunhofer ISI, Prognos AG [Online]’. Available at: https://irees​.de​/wp​-content​/uploads​/2021​/ 03​/Comprehensive​-Assessment​-Heating​-and​-Cooling​_Germany​_2020​.pdf (Accessed 12 January 2022). Transsolar Software Engineering. (2021). ‘TRNSYS: TRaNsient SYstem Simulation Program [Online]’. Available at: https://trnsys​.de​/en (Accessed 9 August 2022). Verein Deutscher Ingenieure. (2012). ‘VDI 2067: Wirtschaftlichkeit gebäudetechnischer Anlagen .beuth​ .de​ /de​ /technische​ Grundlagen und Kostenberechnung [Online]’. Available at: https://www​ -regel​/vdi​-2067​-blatt​-1​/151420393#:~​:text​=Die​%20Richtlinienreihe​%20VDI​%202067​%20behandelt​ ,Richtlinienreihe​%20in​%20mehrere​%20Bl​%C3​%A4tter​%20gegliedert. (Accessed 29 August 2022).

A mixed-integer linear programming approach  363

Weinand, J. M., Kleinebrahm, M., McKenna, R., Mainzer, K., and Fichtner, W. (2019). ‘Developing a combinatorial optimisation approach to design district heating networks based on deep geothermal energy’. Applied Energy, vol. 251, p. 113367. DOI: 10.1016/j.apenergy.2019.113367. Wirtz, M., Kivilip, L., Remmen, P., and Müller, D. (2019). ‘Optimal design and operation of largescale heat pumps in district heating and cooling systems’. In: ECOS 2019 – Proceedings of the 32nd International Conference on Efficiency, Cost, Optimisation, Simulation and Environmental Impact of Energy Systems [Online]. Available at: https://www​.scopus​.com​/inward​/record​.uri​?eid​=2​-s2​.0​ -85079671224​&partnerID​= 40​&md5​=7bb​f94b​5390​95c7​cfd2​54ee​9df6825c6. Wirtz, M., Kivilip, L., Remmen, P., and Müller, D. (2020). ‘5th generation district heating: A novel design approach based on mathematical optimisation’. Applied Energy, vol. 260, p. 114158. Wirtz, M., Neumaier, L., Remmen, P., and Müller, D. (2021). ‘Temperature control in 5th generation district heating and cooling networks: An MILP-based operation optimisation’. Applied Energy, vol. 288, p. 116608. Wolff, D., and Jagnow, K. (2011). ‘Untersuchung von Nah- und Fernwärmenetzen: Überlegungen zu Einsatzgrenzen und zur Gestaltung einer zukünftigen Fern- und Nahwärmeversorgung [Online]’. Available at: https://www​.shk​-thueringen​.de​/de​/fachbetriebe​-finden​/download​/5080​/Studie​%20-​%20 Nah-​%20und​%20Fernw​%C3​%A4rmenetze​%20IWO​%202011​.pdf (Accessed 29 August 2022).

PART VI RENEWABLE ENERGY POLICY

16. The economic analysis of renewable energy policies: a general overview and a historical perspective Christoph P. Kiefer, Pablo del Río and Leticia García-Martínez

INTRODUCTION Renewable energy has been supported for decades now, using different instruments and even different policy approaches. In the 1970s, 1980s and early 1990s, the goal was to encourage the improvement of the technologies through research, development and demonstration (RD&D) support. However, as the technologies became more mature, the emphasis shifted to encouraging their cost reductions through dynamic economies of scale and learning effects as a result of their diffusion. This has been the policy approach until now, and it has been, indeed, quite successful, as several of these technologies (notably on-shore wind, off-shore wind, solar PV and concentrated solar power (CSP)) have experienced impressive cost reductions in the last two decades, in some cases (solar PV and on-shore wind) becoming competitive with other renewables (e.g. hydro) or even with other fossil-fuel technologies on the basis of private or so-called direct costs (see below) alone (IRENA, 2021, 2022). This has led to a discussion on the need to phase out support for these technologies, now or in the near future. The aim of this chapter is to provide a critical, historical overview of the evolution of public policy support for renewable energy technologies, with an emphasis on the last two decades, deployment support and the solar and wind technologies. Accordingly, the chapter is structured as follows. First, the rationale for granting support to these technologies is provided in the next section. Second, a description of those instruments, within the broader set of policy conditions (including targets), is provided and their pros and cons are discussed. Then, we identify the patterns and trends in the adoption of those instruments in the last decades, with a focus on the European Union. A discussion of the alternatives for the design of different instruments is provided. The last section concludes.

WHAT HAS BEEN THE JUSTIFICATION FOR PUBLICLY SUPPORTING THE DEPLOYMENT OF RENEWABLE ELECTRICITY PROJECTS? Humanity faces the major challenge of mitigating climate change caused primarily by the emission of greenhouse gases (GHGs) as a result of burning fossil fuels to generate energy and, in particular, electricity. It is therefore imperative to deepen the process of sustainable energy transition, replacing these fossil energy sources with non-GHG emitting ones. 365

366  Handbook on the economics of renewable energy

STEPS

2040

2019

2040

SDS

SDS STEPS

2019

Generaon (TWh)

Capacity (GW)

The penetration of renewable energy sources for electricity generation is critical in decarbonising electricity systems, a main goal in the context of the energy transition. In this context, renewable energies play an important role. The European Union has agreed on the goal of achieving climate neutrality by 2050, which implies the massive diffusion of renewable energies. Among renewable energy capacity additions, wind electricity and solar PV have played a dominant role in the European Union and elsewhere, fuelled by technology cost reductions and public promotion schemes. There is agreement that wind and solar PV will be the renewables with the greatest potential for global diffusion, given their relatively low costs compared to other renewables and even fossil fuels. According to the IEA (2021), renewable electricity generation will increase by between 2.6 and 3.9 times in the 2019–2040 period (depending on the scenario) and renewable energy capacity will increase by between 3.9 and 4.3 times in the same period. The share of renewables in electricity generation will increase from 27% in 2019 to between 47% and 72% in 2040 (depending on the scenario). This increase will be triggered by solar PV and on-shore wind. Figure 16.1 shows the trends of these two technologies in the considered period. At the EU level, the promotion of renewable electricity has been a policy goal not only in the most obvious fields of climate, energy and environmental policy but also in other policy fields such as regional and rural development policy and employment policy as well. Justification for granting support to renewable electricity at both the EU and Member States (MS) levels has been, at least, twofold. On the one hand, the benefits from renewable energy are taken

0

2000

4000

World-PV

6000

8000

10000

World-wind

Note:   in the STEPS scenario (The Stated Policies Scenario), COVID-19 is gradually brought under control in 2021 and the global economy returns to pre-crisis levels in that year. This scenario reflects all of today’s announced policy intentions and targets, insofar as they are backed up by detailed measures for their realisation. In the Sustainable Development Scenario (SDS), a surge in clean energy policies and investment puts the energy system on track to achieve sustainable energy objectives in full, including the Paris Agreement, energy access and air quality goals. The assumptions on public health and the economy are the same as in the STEPS. Source:   own elaboration with data from IEA (2021)

Figure 16.1  Solar PV and on-shore wind installed capacity and generation in 2040 in the world compared to 2019

The economic analysis of renewable energy policies  367

into account. Electricity from renewable energy sources (RES-E) pollutes much less than conventional electricity and, therefore, avoids its negative environmental externalities. More specifically, RES-E contributes to the achievement of the climate change mitigation goals, established internationally by the Kyoto Protocol (in the past) and the Paris Agreement (for the future). However, climate change mitigation is only one of the possible benefits from RES-E deployment. Emissions of local pollutants are reduced as well. Apart from the environmental benefits, RES-E provides other socioeconomic advantages such as development, employment and investment opportunities (see Chapter 5). Finally, by having renewable energy projects in its territory, Europe can also reduce its fossil-fuel dependency and mitigate the risks related to the security of energy supply, which is certainly a major policy concern nowadays. Some efforts to calculate the socioeconomic and environmental benefits of CO2 emissions reductions, the avoidance of fossil fuels and employment creation due to RES-E deployment have been made (see also Chapter 5). For example, Ortega-Izquierdo and del Río (2020) assess those socioeconomic and environmental benefits of wind energy deployment in the European Union and their evolution in the period 2008–2016. Their results show that, overall, those benefits are considerable and have increased over time. However, some benefits are more relevant than others. In fact, in monetary terms, fossil-fuel savings are twice the benefits of CO2 emissions abatement. Their relevance also differs across countries and, for employment creation, across distinct stages of the value chain. However, those benefits stemming from renewable electricity would not be enough by themselves to justify the implementation of policy support schemes because, if the market valued such advantages (avoided negative externalities), then RES-E would penetrate the market without any need for public support. Unfortunately, this has not been the case, since renewable electricity has competed on an unequal playing field with thermal-based conventional electricity (del Río and Gual, 2004). From a societal point of view, the focus should not (only) be on the private costs, but on social costs, which are the result of adding the external costs (the avoided negative environmental externalities due to burning fossil fuels) to those private costs (which can be measured with the levelised cost of electricity, or LCOE) (see also Chapter 2). Although, in general, private generation costs have been higher for renewable than for conventional electricity, the former provides benefits that are not valued by the market. Those benefits translate into a generally lower social cost (inclusive of private costs plus negative external costs minus positive externalities) for renewable energy, but market operators (investors, generators, suppliers and consumers) are guided by the incentives provided by the market, where decisions are taken on the basis of private and not social costs (unless, of course, policy measures internalise these externalities) (del Río and Gual, 2004). Some research efforts have been made in the past to compare the private costs and environmental externalities of each electricity source and, thus, the social costs of different energy sources to generate electricity. For example, Harmon and Cowan (2009) showed that the private costs of renewable electricity generation technologies were, almost two decades ago, much higher than the private costs of conventional technologies. However, when the negative environmental externalities of different sources were considered, the social costs were more balanced between the different technologies (Table 16.1). Therefore, investment decisions in electricity generation in general and renewables in particular should be made taking into account the social costs of the different energy generation sources that depend, in turn, on public policy internalising these externalities. This is why their role in the energy transition is so important.

368  Handbook on the economics of renewable energy

Table 16.1  Private and social generation costs for different renewable energy technologies (in c$2006/kWh) Generation costs (private) (1)

External costs (2)

Generation costs (social) (1) + (2)

Solar

21.5–48.0

0.7

22.2–48.7

Wind

4.0–6.0

0.1–0.3

4.1–6.3

Biomass

7.0–9.0

0.2–3.4

7.2–12.4

Hydro

2.4–7.7

0.0–1.1

2.4–8.8

Natural gas

3.9–5.0

1.1–4.5

5.0–9.5

Coal

4.8–5.5

2.3–16.9

7.1–22.4

Nuclear

11.1–14.5

0.2–0.8

11.3–15.3

Source:   Harmon and Cowan (2009)

Public support for renewable energy levels the playing field with respect to conventional electricity, internalises the positive externalities of renewable energy in the decisions taken by economic actors and allows renewable energy to penetrate the electricity market. Therefore, the justification for supporting renewable electricity in general has traditionally been based on the dichotomy of its lower social costs but higher private costs, although the situation has changed recently (see the last section of this chapter). Moreover, in the past, an important argument in support of renewables has been related to cost dynamics. The diffusion of a technology leads to cost reductions and technological improvements due to learning effects and dynamic economies of scale. Learning effects occur continuously because, during the process of manufacturing the technology, or during its diffusion, improvements in the technology occur as a result of suggestions from users (learning by using) or because manufacturers themselves realise that there are better ways of producing the technology (learning by doing). These effects give rise to a positive externality (Stern, 2007): even those firms that did not initially invest in the new technologies can benefit from these cost reductions. The initial investor cannot capture these learning benefits. Therefore, investments in the new technology will be below socially optimal levels. Such learning effects are the result of previous investments and, therefore, there is circularity: high-cost technologies do not diffuse as a result of these high costs, but they have high costs precisely because they do not diffuse. To start this diffusion process and break the vicious circle, public policies are needed to stimulate the adoption of technologies when they are still expensive. When a dynamic approach is considered, promoting renewable energy is even more justified. In a situation of technology competition such that in the electricity market, the dominant technology (conventional electricity) has been able, through learning effects, to reduce its costs and to be better known by economic actors. This creates a feedback effect: this technology is adopted more and more because it has been technically improved and it is cheaper. Emerging renewable energy technologies do not even have a chance in this situation unless, of course, they are initially supported. However, while the justification for supporting renewable electricity in general has traditionally been based on the dichotomy of lower social costs but higher private costs, the situation changed some years ago. Two main renewable energy technologies for electricity generation (PV and wind) are competitive with respect to their fossil-fuel counterparts in

The economic analysis of renewable energy policies  369

terms of private costs (as measured by their LCOEs), which calls into question the need for further public support for them. This issue is addressed in the fourth section of this chapter. It is important to note that the costs of renewables are not only the private (direct) costs but also the costs to the system, the sum of the direct costs (mentioned above) and the indirect costs. Direct costs refer to investment and operation and maintenance (O&M) costs, and they are usually measured as LCOE. Indirect costs include the adjustment service cost (deviations of actual generation from scheduled generation (rolling reserve and intraday adjustments to ensure system stability)), profile costs (including back-up costs) and grid costs (grid extension and reinforcement) (see Chapter 2 for further details). With respect to direct costs, these indirect costs are low with low levels of penetration for non-manageable renewables: between 1% and 16% (Gowrisankaran et  al., 2011; Gross et  al., 2006; Holttinen, 2011; Khatib and Difiglio, 2016; Kopsakangas-Savolainen and Svento, 2013). However, with increasing levels of penetration of non-dispatchable renewables, integration costs will probably increase more than proportionally. Several authors show that back-up costs for renewables tend to increase in direct proportion to the renewable capacity (MW). The costs of balancing services increase because more rolling reserve capacity is required at low load factors. In turn, grid extension and reinforcement costs increase because the grid must be extended to find suitable locations for renewable projects and must be reinforced to carry more load (Edenhofer et  al., 2013; Khatib and Difiglio, 2016; Stram, 2016).

SETTING THE SCENE: THE COMPONENTS OF POLICIES FOR THE SUPPORT OF RENEWABLE ELECTRICITY Too often, “policies” have been equated with “instruments”. However, in reality, these are different terms (see Rogge and Reichardt, 2016, for details).1 We adopt a pragmatic and hierarchical view of policies, which would be made up of top-down components, such as framework conditions (including targets and policy stability), but also components with a higher degree of granularity, including instruments, but also how these are designed (the choices of design elements within particular instruments). All these components, which are described below, would have a strong influence on the attractiveness of investing in RES-E projects for investors. Framework Conditions: Targets and Policy Stability The first thing to bear in mind is that, beyond the specific instrument chosen to promote wind energy, it must be implemented with appropriate framework conditions in order to attract RES-E investments. There are at least two crucial aspects of these framework conditions: the existence of long-term objectives and regulatory stability. Specific instruments should be included within appropriate policy frameworks which, in turn, should be credible and stable, with long-term visions and targets. A long-term horizon for energy policies is needed to provide a signal for investors of the existence of a future market which, thus, attracts investments. Therefore, companies and consumers would have confidence that investments in climate-friendly technologies will eventually pay off (Groenenberg and de Coninck, 2008). The transition to a low-carbon economy offers opportunities for new products and services, and if policy can paint a credible picture of this vision, then this can

370  Handbook on the economics of renewable energy

attract the desired investment. Visions are particularly helpful to guide the multitude of decisions that are to be made by different actors (del Río, 2011; Sandén and Azar, 2005). On the other hand, policy stability is often cited as a critical enabler of investments in lowcarbon technologies such as renewable energy technologies (del Río et al., 2012b). A country where policies are constantly changing, especially retroactively or retrospectively, is unattractive for investments in renewable energy deployment (del Río et al., 2021). However, stability should not mean policy rigidity, i.e. that some changes in certain design elements cannot be modified in response to changing circumstances or to detected drawbacks of the policies (policy learning). Thus, some degree of policy flexibility should be allowed, but no sudden “change in direction” should be imposed (Reichardt et al., 2017). Instruments: Carbon Prices, R&D Support and Deployment Support Pricing carbon A price on the carbon content of fuels used in electricity generation is a classic instrument that internalises the externalities of CO2 emissions. Such a price, which can be set through taxes (CO2 tax) or through emissions trading systems (ETS), like the one in place in the EU since 2005, incentivises investments in low-carbon technologies in general and renewables in particular. Unfortunately, a carbon price may not be effective in promoting the diffusion of less mature renewable technologies and must be high enough to promote mature ones with high relative costs (which usually makes it politically unfeasible). In the EU ETS, this price has recently been at such high levels (above €50/tonne CO2 since May 2021 and around €80/ tonne CO2 during 2022), but this has not been the case for the rest of the system’s lifetime, with constant fluctuations in this price and levels below €10/tonne CO2 for most of that period (until 2018). It is unlikely that these levels would have encouraged investment in renewables. It is therefore necessary to complement it with instruments that specifically support renewable technologies, through support for research, development, demonstration and deployment. Supporting research, development and demonstration (RD&D) The economics of innovation shows that the role of governments is most effective when combining “supply-push” instruments (which influence the supply of knowledge in a given technology) with “demand-pull” instruments (which support diffusion). Supply-push instruments involve directly promoting technological development, through R&D support. Several types of instruments can be used to support R&D in renewables in general: directly through R&D funding in public research centres or universities, or indirectly through fiscal measures aimed at promoting R&D investments by companies, such as subsidies or tax breaks. In addition, demonstration can be encouraged through the funding of demonstration projects. Demandpull instruments also include support for education and training and the promotion of knowledge networks (Mir-Artigues et al., 2019).2 Deployment Support As mentioned above, technology diffusion leads to cost reductions and technological improvements due to learning effects and dynamic economies of scale, and public policies are needed to stimulate the adoption of technologies when they still have high costs. The following section describes and analyses the different instruments for the diffusion of renewable technologies.

The economic analysis of renewable energy policies  371

A crucial current issue is whether deployment support is needed for renewable energy technologies. It can certainly be justified according to the aforementioned arguments for less mature technologies, for which a cost-gap exists. However, those arguments are less suitable for renewable energy technologies which have already advanced along their learning curves, with relatively less potential for further cost reductions, and which are already competitive on the basis of private costs with their fossil-fuel competitors. This is the case with solar PV and on-shore wind. According to IRENA (2021, p. 11), the cost of electricity from solar and wind power has fallen to very low levels. Since 2010, globally, a cumulative total of 644 GW of renewable power generation capacity has been added with estimated costs that have been lower than the cheapest fossil fuel-fired option in each respective year. In emerging economies, the 534 GW added at costs lower than fossil fuels, will reduce electricity generation costs by up to USD 32 billion this year. New solar and wind projects are increasingly undercutting even the cheapest and least sustainable of existing coal-fired power plants. IRENA analysis suggests 800 GW of existing coal-fired capacity has operating costs higher than new utilityscale solar PV and onshore wind, including USD 0.005/kWh for integration costs.

Table 16.2 shows the costs of renewable energy technologies in 2020. According to IRENA (2021), the costs for electricity generation technologies based on fossil fuels are in the range of 0.175 to 0.05 $/kWh. Therefore, it can be argued that solar PV, on-shore wind and hydro are already cost-competitive technologies in terms of direct costs (LCOE) with respect to other renewable and non-renewable electricity generation technologies. Why should support be provided in this case, when their relative costs suggest that market forces would lead to their massive diffusion? There might be several reasons for this: (1) the range of costs is global (so, in some places, wind and PV may not be competitive); (2) their high capital intensity and, thus, need for large up-front investments make it necessary to provide a floor price on the remuneration (which, otherwise, would only come from the electricity market price) in order to ease financing for these projects; (3) wholesale electricity prices are likely to go down (although this is a debatable issue) due to the increasing and large penetration of low-variable cost electricity generation technologies, such as renewables. This Table 16.2  Global weighted-average LCOE for different renewable energy technologies in 2010 and 2020 Electricity generation technologies

LCOE ($/kWh)* 2010

2021

Biomass

0,076

0,078

Geothermal

0,049

0,068

Hydro

0,038

0,048

Solar PV**

0,381

0,048

CSP

0,340

0,114

Off-shore wind

0,162

0,075

On-shore wind

0,089

0,033

Notes:   * 2020 $/kWh ** Large-scale PV Source:   IRENA (2021)

372  Handbook on the economics of renewable energy

“price-cannibalisation effect” would push electricity prices down and would make the financing of renewable energy projects difficult; (4) the ambitious renewable energy targets mean that not only the cheapest technologies and locations will be needed to comply with them, but also more expensive ones, which means that remuneration based only on wholesale prices may not be sufficient.

WHAT DEPLOYMENT INSTRUMENTS ARE OUT THERE AND WHAT ARE THEIR PROS AND CONS? Deployment Instruments Several alternative instruments have been used in the past to promote the deployment of renewable electricity projects. They can be grouped according to different classifications. Most of them support generation, but others support the deployment of capacity (investment subsidies and soft loans). In price-based instruments, the level of support is set by the government (administratively‑set remuneration) and the market operators (investors) decide on the quantity that they want to invest in. In quantity-based instruments, the government sets a total amount of capacity (or electricity) that it wants to support and it is the interaction between the supply and the demand sides of the market which lead to a given level of support (e.g., as in auctions and quotas with tradeable green certificates). This last classification has become a crucial one from a historical perspective. The reason is that, as explained in the next section, price-based instruments may lead to a lack of control for the RES capacity included in the system and, thus, higher total costs of support. This concern has led governments to replace them with quantity-based schemes in recent years. Promotion has traditionally been based on three main (primary) mechanisms: one pricebased instrument (administratively set feed-in tariffs or feed-in premiums), and two quantitybased ones (tradable green certificates (TGCs) and bidding/tendering systems). These have been supplemented by other complementary instruments (investment subsidies, fiscal and financial incentives and green pricing). Countries usually apply one (or, at most, two) of the schemes in the first group and complement it with other measures pertaining to the second group. Instruments may be differentiated per technology. In all these instruments, the renewable generator receives the support, but the cost of the support is ultimately borne by the electricity consumer. Funding of support through the public budget is much less common. A brief description of those instruments is provided below (del Río and Gual, 2004; Mir-Artigues and del Río, 2016). ●

Feed-in laws provide for preferential prices per kWh (or MWh) of RES-E generated, paid in the form of guaranteed premium prices and combined with a purchase obligation by the utilities. Feed-in tariffs (FITs) provide total payments per kWh of electricity of renewable origin, whereas a payment per kWh on top of the electricity wholesale-market price is granted under feed-in premiums (FIPs).3 In FITs, the electricity generator receives full support for its electricity. It is independent of the price of electricity on the wholesale market and, therefore, generators do not have an incentive to sell the electricity at those times when it is most valuable to the system (i.e. when electricity demand is highest). This is not the case with FIPs. In this case, what is granted to the renewable generator is an amount

The economic analysis of renewable energy policies  373

●

●

of support (also per kWh) that is additional to the revenue it obtains from the sale of the electricity in that market. As the price of electricity varies according to electricity demand levels, and is higher when electricity demand is higher, the renewable generator will have an incentive to sell electricity at precisely those times when electricity is most valuable. FIPs therefore facilitate the integration of renewables into the electricity market by providing that price signal (to which all other electricity generation technologies are also subject) and are more compatible with a liberalised electricity market. However, with respect to FITs, and as a consequence of this greater exposure to the market, FIPs provide less security for the generator/investor (which usually translates into higher risk premiums and more expensive financing with respect to FITs), as the remuneration has a non-predictable component (the market price of electricity itself). In other words, there is a clear trade-off or conflict between incentives for market integration and security for the investor. TGCs are certificates that can be sold in the market, allowing RES-E generators to obtain revenue. This is additional to the revenue from their sales of electricity fed into the grid. Therefore, RES-E generators benefit from two streams of revenue from two different markets: the market price of electricity plus the market price of TGCs multiplied by the number of MWh of renewable electricity fed into the grid. The issuing (supply) of TGCs takes place for every MWh of RES-E, while demand generally originates from an obligation, usually on electricity suppliers. Electricity distribution companies must surrender a number of TGCs as a share of their annual sales. Otherwise, they have to pay a penalty. The TGC price covers the gap between the marginal cost of renewable electricity generation at the quota level and the price of electricity. Auctions. The government invites RES-E generators to compete for either a certain financial budget or a certain RES-E generation capacity. Defined as technologically neutral or within a given technology band, the cheapest bids per kWh are awarded contracts and receive the subsidy. The operator is paid the bid price or auction clearing price per kWh. Note that the levels of FITs or FIPs can either be set administratively (through a decision of the government) or through auctions. Therefore, it is not correct to say that there is a dichotomy between FITs/FIPs and auctions.

Several secondary instruments have been combined with the former in the past, including: ●

●

●

Investment subsidies. These are granted at the beginning of the project lifetime and can be calculated as a percentage of the renewable energy output or the specific investment cost, although this latter version is more common. Investment grants for RES-E are available in many EU member states. Soft loans are usually provided by governments at a rate below the market interest rate. Low-interest loans can be offered by the government directly through state-owned banks or through subsidies to commercial banks. In some cases, they can significantly reduce the costs of capital. Soft loans may also provide longer repayment periods or interest holidays. In short, they involve more favourable conditions for borrowers. Fiscal incentives can be exemptions or rebates on (energy, corporate or income) taxes, tax refunds, lower VAT rates or attractive depreciation schemes.

In contrast to primary instruments, which generally cover all RES-E installations and are set at the national level, secondary instruments are usually limited in scope and

374  Handbook on the economics of renewable energy

circumscribed to specific types of projects (e.g., small ones) and technologies (e.g., solar PV). Whereas main instruments are almost always applied at the national level, secondary ones are often applied at both the national and lower government levels, that is, regional/ provincial/municipal. Main Criteria to Assess the Instruments The success of the functioning of any RES-E support scheme can be judged against several criteria (Table 16.3).4 Note that the minimisation of generation costs (static efficiency) and support costs are not totally independent of each other (Cerdá and del Río, 2015). Cost-containment is a goal that is related to but distinct from static efficiency, which often depends on the ability of a policy scheme to limit capacity growth, regardless of the unitary (€/MWh) cost. It refers to the support that is paid to RES-E generators and which usually falls on electricity consumers. In fact, both are related by “investors’ risks” which in turn are influenced by the type of RES-E support policy. An effective and cost-efficient RES-E policy is risk-conscious and does not introduce unnecessary policy-related risks. Low cost for loans and equity would reduce the cost of RES-E projects and the required financial support from governments or consumers, while more investments into RES-E projects can be attracted and more RES-E projects can be realised (i.e., the system becomes more effective). A country with RES-E policies leading to lower investors’ risks will experience more RES-E growth at lower specific generation costs. Lower generation cost can be translated almost 1:1 to lower required support policy cost (Rathmann, 2011). Both criteria also interact in more indirect ways. For example, large RES-E support costs may trigger social rejection and lead to retroactive changes which, in turn, increase the risk of investing in such a country and, thus, the costs of capital and the investment costs, leading to higher generation costs. Assessing the Pros and Cons of Different Instruments Minimisation of generation costs This criterion refers to the minimisation of (system) generation costs, which are the sum of the direct and indirect costs of electricity generation. In general, given the considerable pressures to reduce costs, it can be expected that quantity-based instruments encourage the placement of RES-E projects in the best locations, i.e., in those places with the best renewable energy resources (solar irradiation, wind, etc.), in order to be more competitive. However, this incentive is not lost with price-based support schemes such as administratively set FITs or FIPs (ASFITs/FIPs), since the projects in better locations lead to a greater profit margin (greater revenues). An important element of the direct (or “private”) costs are the financing costs. Among other factors, these depend on the risks for investors under different schemes. In principle, pricebased schemes lead to greater revenue certainty and, thus, lower risks. This revenue certainty can also be achieved through auctions, but in this case there is a risk of not being awarded (therefore losing the costs incurred before). The lower investors’ risks under price-based schemes suggest a lower risk premium for borrowers and lenders and, thus, better financing conditions, i.e., lower financing costs.

The economic analysis of renewable energy policies  375

Table 16.3   Assessment criteria Criteria

Description

Indicators (examples)

Effectiveness

Degree to which the RES-E support scheme results in deployment of RES-E projects.

Realisation rate (%).

Static efficiency (cost-effectiveness)

Reaching the target at the lowest possible overall generation costs. The outcome of a RES-E support instrument is efficient if the RES-E target is achieved at the lowest system costs (i.e., addition of direct and indirect costs). An important aspect here is market integration – the compatibility of the policy with the principles of market integration, which may include electricity market exposure or balancing requirements.

Total generation costs (system costs) (€, €/MWh). These are the sum of the direct costs (LCOE) and the indirect costs.

Dynamic efficiency

This refers to long-term technology effects, including impact on innovation, technology diversity and cost reductions over time.

Private R&D investments (€) and number of patents. Evolution of the share of different technologies over time (%). Evolution of the costs of the technologies over time (€/ MWh).

Minimisation of support costs

Impact on the level of support for different technologies (average and total). This is usually paid by electricity consumers.

Average support level per technology (net of generation costs) (€/MWh). Total support costs net of total generation costs (€).

Local impacts

Impact on several variables at the national, regional and local levels. They can be environmental or socioeconomic and include emissions of GHG and local pollutants, variations in fossil-fuel energy dependence, employment effects, industry creation, regional development and export opportunities.

GHG emissions being reduced (additional to the ETS) (tonnes). Emissions of local pollutants reduced (tonnes). Reduction of fossil-fuel imports: trade balance affected (avoided fossil-fuel consumption). Local content/promotion of local industry. Regional concentration of deployment. Additional jobs in the renewable energy sector (number). (Continued)

376  Handbook on the economics of renewable energy

Table 16.3  (Continued) Criteria

Description

Indicators (examples)

Sociopolitical feasibility

Degree to which the design elements and the whole support scheme are socially acceptable and politically feasible. This depends on other criteria (minimisation of support costs, the existence of positive and negative local impacts from RES-E deployment, etc.). A main aspect is whether the design element or support scheme fits into the existing institutional structure.

Fit to decision-makers’ institutional capacity. Qualitative variable (more/ less acceptable; more/less politically feasible).

Legal feasibility

Extent to which a given design element or the whole support scheme complies with EU legislation (primary and secondary law), including state aid rules and internal market principles.

Compliance with state aid rules (Y/N). Compliance with internal market principles (Y/N).

Source:   del Río (2017)

On the other hand, whether the support scheme leads to higher or lower indirect costs represents a main criterion by which to judge its pros and cons. More specifically, whether support schemes favour the integration of RES-E into the electricity market is a main aspect. In this case, TGC schemes and fixed FIPs, whether administratively set or set in auctions, can be deemed superior to other instruments. The reason is that, in these schemes, RES-E generators have to sell their electricity in the wholesale market and receive a remuneration for this (the price of electricity in this wholesale market, in €/MWh). In addition, they receive a complementary remuneration, also in €/MWh, either the TGC price (under a TGC scheme) or a premium (under a fixed FIP). As mentioned before, in both cases, they have an incentive to sell their electricity at times of greater electricity price levels (when the demand is high), which is also when the value of electricity is greater, i.e., when the RES-E electricity fed into the grid has the highest value for the electricity system. They are therefore considered beneficial to market integration. This is not the case under FITs (whether set administratively or in auctions) since, in this case, there is not a double source of remuneration, but only one source of remuneration, which is detached from the moment when or the place where the electricity is delivered, i.e., FITs do not support RES-E integration when RES-E has more value to the electricity system. The above discussion suggests that there might be a trade-off between the different components of the direct and indirect costs of electricity generation. Minimisation of support costs Support for RES-E projects is finally paid by either electricity consumers in their bills (the usual case) or by taxpayers though the public budget. Two relevant aspects in this context are the unitary costs of support (the difference between the level of support and either the costs of the technologies or the wholesale electricity price) and the total costs of support, which are the result of multiplying the unitary support by the total amount of RES-E which is eligible for support. The former is measured as €/MWh, whereas the latter is measured in million euros.

The economic analysis of renewable energy policies  377

ASFITs have traditionally been identified as problematic in terms of support cost-containment. For example, the impact assessment accompanying the document “Communication from the Commission Guidelines on State Aid for Environmental Protection and Energy for 2014–2020” states that “administratively established support levels do not ensure cost-efficiency due to the information asymmetries between the regulator establishing the support level and the producers that benefit” (p. 18). However, the empirical evidence suggests that price-based support schemes do not necessarily score worse than quantity-based schemes regarding the unitary support costs. Rather, the opposite seems to have been the case in the past, at least for on-shore wind in European Union member states (see, e.g., Ragwitz et al., 2012). The reason is that ASFITs for this technology have been set at rather reasonable levels, i.e., not high above its generation costs. Of course, there is always a risk that the aforementioned asymmetric problem occurs when setting support levels in price-based schemes, especially for less mature technologies, which may lead to high unitary support costs.5 However, the risk in quantity-based ones is that the level of support, which is either set by the TGC price (in TGC schemes) or by the marginal bid (in uniform-pricing auctions), leads to a too high level of unitary support for the cheapest technologies and/or locations (e.g., an excessive remuneration for mature technologies). In addition, a high risk premium resulting from the uncertain development of the electricity and the certificate price typically increases policy costs. This seems to be the reason for the aforementioned result of wind energy in the EU. However, the total amount of support is as important as the unitary level of support. Both are usually related in price-based schemes without a cap on the capacity which is eligible for support. The reason is that, in this case, a high level of unitary support encourages investments in RES-E, which leads to a high level of total support (remember, unitary support times the generation of electricity which is eligible for support) unless such a cap exists. In the past, for example, this led to the solar PV booms and, thus, to a large increase in total support costs (see del Río and Mir-Artigues, 2012; Mir-Artigues and del Río, 2016). This uncertainty on total support costs if large RES-E volumes are deployed reduces the attractiveness of price-based support schemes for policy makers. High levels of support would reduce the social acceptability of renewables or renewable energy support. Of course, this risk could be limited by including a cap on the amount of generation which is eligible for support, but the adoption of this design element was rare in past price-based schemes. In contrast, capacity or generation caps are an inherent feature of TGC and auction schemes, which limit such risk. There seems to be a clear trade-off between the minimisation of the risks for investors (of volatile prices and, thus, revenues) and minimising the risks for consumers (due to large total support costs). Under price-based support schemes, the risks for investors have been lower, but total support costs, which fall on consumers, have been higher. Effectiveness Effectiveness refers to the realisation rates of projects and has been deemed relatively higher in price-based schemes compared to quantity-based ones. ASFITs have proven to be quite effective in triggering RES-E deployment in the past (del Río, 2008; del Río and Mir-Artigues, 2012), more effective than TGC schemes (Ragwitz et al., 2012). The reason is probably related to the aforementioned greater certainty of remuneration under ASFIT/FIPS. In contrast, the TGC price may be too volatile under TGC schemes, which may make RES-E investments less

378  Handbook on the economics of renewable energy

attractive for potential investors. In addition, banks may be less eager to fund these projects, given such risks on revenue streams. As mentioned above, FIPs or FITs may be set administratively or in auctions. Both would lead to secure revenues. Therefore, certainty of the remuneration cannot be the distinct feature between them. Delays in building the projects and project cancelations have been frequent in the past (del Río and Linares, 2014; IRENA, 2019). Dynamic efficiency A crucial aspect for some governments, at least in Europe, is whether the support scheme encourages innovation, cost reductions and a diversity of technologies. Dynamic efficiency refers to the ability of an instrument to generate a continuous incentive for technical improvements and costs reductions in renewable energy technologies: i.e. an incentive to positively influence technological change processes in the medium and long term. Those RES-E support instruments which favour the commercialisation of expensive technologies in niches tend to lead to quality improvements and cost reductions; this will allow us to have renewable energy technologies in the future to comply with more ambitious renewable energy and emissions reduction targets at reasonable costs. If currently expensive mitigation technologies have a large cost reduction potential with increased diffusion (as shown by several studies for energy technologies), then supporting them today would lead to welfare benefits in terms of intertemporal mitigation efficiency (i.e. cost-effectiveness in the short, medium and long term) (del Río, 2017; del Río et al., 2015a). In contrast to static efficiency, dynamic efficiency has an intertemporal perspective on costs. The impact of RES-E support schemes upon innovation in renewable energy technologies has several aspects or “dimensions”: diversity, R&D, learning effects and competition (see del Río and Kiefer, 2022). Diversity is about supporting different technologies, but also different actors. Some authors have claimed that vested interests are a barrier to a transition to renewable energy technology systems. New energy technologies are often developed outside the established energy systems and engage non-traditional energy actors (Åstrand and Neij, 2006). Actors, networks and institutions involved in radical innovation processes are not identical to those performing activities that sustain an established system (Markard and Truffer, 2008). In general, it can be argued that the more neutral the support scheme, the greater the level of static efficiency and the lower the level of diversity (of technologies, project sizes, locations and actors), although not necessarily the lower the support costs. In fact, technology-neutral instruments such as quotas with TGCs have led to excessive remuneration for some technologies, as reported above. The higher the level of effectiveness in creating a market for renewable energy technologies, the higher the level of dynamic efficiency. In this context, ASFITs have been more effective than other instruments (i.e., quotas with TGCs) in creating a market for different types of renewable energy technologies and increasing the deployment of RES-E projects. This has sometimes had a negative side in terms of an uncontrolled increase in deployment for some technologies and, thus, the skyrocketing of support costs, especially for very dynamic technologies experiencing substantial cost reductions over time (such as solar PV). But, on the other hand, a higher degree of effectiveness has positive impacts on several criteria, notably dynamic efficiency and local impacts (benefits). ASFITs have resulted in strong domestic industries in several countries (Mitchell et al., 2011). The creation of a market for the technologies has led to technological diversity and innovation, i.e. a greater level of dynamic efficiency (del Río et al., 2012a, 2015b). ASFITs may be particularly

The economic analysis of renewable energy policies  379

suited to supporting less mature technologies or small-scale applications, which have difficulties in bearing the price risks or the transaction costs for participation in a market platform with professional traders (Held et al., 2014). Furthermore, the instrument has been considered particularly suitable to encourage a diversity of actors compared to other instruments. This is probably related to the lower risk, the creation of a space (market) for different types of technologies and investors and the simplicity of the instrument. As argued by del Río et al. (2012a) and Mitchell et al. (2011, p. 58), which review an abundant literature on the topic, ASFITs tend to favour ease of entry, local ownership and control of RES systems and thus can result in wider public support for RES. Mendonça et al. (2009) found that steady, sustainable growth of RE would require policies that ensure diverse ownership structures and broad support for RES. The local benefits have made this instrument quite attractive for policy makers willing to support renewable energy technologies and to develop a supply chain for them. In turn, their social acceptability and political feasibility can be deemed high in the initial stages of deployment, but the skyrocketing of support costs for some technologies in some countries has reduced their political feasibility, leading to retroactive cuts in some countries. Furthermore, the EU State Guidelines call their legal feasibility into question, although they leave the door open for their use under given circumstances. In contrast, quotas with TGCs have generally been less effective than FITs in creating a market for renewable energy technologies, especially so for the less mature technologies, for which quotas with TGCs have hardly been effective. They have not promoted a diversity of technologies where they have been implemented (Mitchell et al., 2011) and, thus, their score on the dynamic efficiency criterion is deemed poor. In addition, they tend to lead to the concentration of projects in certain locations where the resource (e.g., wind) is most abundant in an attempt to maximise their income. This is an outcome of the uncertainty attached to TGC prices and may increase the likelihood of social acceptance problems due to “not-inmy-backyard” (NIMBY) phenomena. Finally, they score poorly regarding actors’ diversity. As stressed by the European Commission (2013, p. 10), they limit “provision of renewables only to large scale incumbents capable of ‘on balance sheet financing’, or with access to cheaper debt financing”. The high risks in the quota obligation system tend to favour incumbent players, since large companies are usually better able to hedge the prevailing price risks (Held et al., 2014). Jacobsson et al. (2009) and Verbruggen and Lauber (2009) argue that it is primarily incumbent actors who would benefit from the new market. The transaction and administrative costs of a TGC system are higher than with FIT, making the participation of small-scale new entrants cumbersome and therefore limited. Their social acceptability and political feasibility are deemed uncertain. On the one hand, low effectiveness and high unitary support costs may be detrimental in this regard. On the other, the relatively low total support costs (given its comparatively low effectiveness) could make it attractive for policy makers. ASFIPs seem to be in an intermediate position compared to ASFITs and quotas with TGCs. Their dynamic efficiency can be expected to be greater than in quotas with TGCs, since they provide greater certainty to investors and are more likely to facilitate the creation of a market for renewable energy technologies. But they are probably less dynamically efficient than ASFITs (given the higher risks and lower market creation compared to this instrument). The diversity of actors being promoted is also likely to be intermediate. The lower risks of ASFIPs and greater simplicity of the instrument compared to quotas with TGCs is likely to encourage a greater diversity of actors. But the risks in ASFIPs are higher than in ASFITs and,

380  Handbook on the economics of renewable energy

in addition, generators have to sell their electricity in the market. Both factors are likely to discourage small actors. Risks fall asymmetrically on different actors, with the smaller ones being less capable of hedging them. And finding a trader in the electricity market is also likely to result in higher unitary transaction costs for the smaller investors/generators. However, the fact that FIPs (whether set administratively or in auctions) are regarded as more “marketcompatible” makes them more legally feasible, at least in Europe. To sum up, it could be argued that both price-based and quantity-based support have their pros and cons. Price-based support in the form of ASFITs/FIPs has proven more useful to kickstart renewable energy sectors and markets than quantity-based schemes in the form of auctions or TGC schemes. They have implied lower risks for investors, have allowed the creation of value chains and have encouraged actor diversity. However, they face the asymmetric information problem, which can be a source of inefficiencies. It has been difficult to control the volume of RES-E investments, capacity and generation and may have led to “booms”. Furthermore, the lower competitive pressure inflicted on actors throughout the whole value chain may induce lower levels of innovation, although this is a debatable issue (see del Río and Kiefer, 2022, for an in-depth discussion). The above suggests that different instruments may be more suitable for different levels of maturity and costs of RETs. Kitzing et  al. (2018) provide an interesting contribution in this regard. Taking into account the evolving risks of technologies throughout their life cycle, the authors argue that the optimal policy choice between price and quantity instruments changes over time. The authors claim that price instruments would be better in the first stages of the technology (early market deployment), because they stabilise revenues and decrease market risks for investors, which accelerates deployment without necessarily compromising economic efficiency. However, in latter stages, a shift to quantity-based schemes is recommendable. Protective policies that work well for niche technologies should, however, be used cautiously during market upscaling and diffusion, due to the changing nature of risks. We use theoretic arguments and a case to demonstrate that a gradual shift towards quantity control may become preferable for welfare maximization under certain circumstances. (Kitzing et al., 2018, p. 369)

Secondary instruments such as investment subsidies, fiscal incentives and soft loans also have their advantages and drawbacks. Investment subsidies (direct grants), which are direct grants provided to project developers when the installation is built, are straightforward and easy for developers to value. They lead to lower capital costs for investors and also allow policy makers to identify whether they have been effective in triggering a given RES-E capacity, since it is not necessary to wait until the end of the lifetime of the plant to identify their effects. However, they do not encourage the plant to be as efficient as possible, since they are received independently of the amount of electricity that the plant produces. Therefore, they do not provide appropriate incentives to encourage strong production, undermining cost effectiveness (del Río, 2017). However, although they create a lower incentive to operate the plant as efficiently as possible, compared to generation-based support, the incentive is not totally lost, since the plant sells its electricity into the wholesale market and, thus, its revenues are greater the higher the efficiency of the plant (amount of MWh produced per MW installed).

The economic analysis of renewable energy policies  381

Fiscal incentives are highly flexible policy tools that can be targeted to encourage specific renewable energy technologies. Soft loans reduce investors’ risks and, thus, capital costs, which is beneficial for policy costs, generation costs and effectiveness. They may be useful for specific technologies and actors and, in particular, smaller actors which have difficulties in accessing credit. However, none of them is deemed an appropriate primary instrument (del Río, 2017).

HISTORICAL TRENDS: WHAT DEPLOYMENT INSTRUMENTS HAVE BEEN USED IN THE PAST? WHICH ONES ARE USED NOW? WHICH WILL BE USED IN THE FUTURE? Using different data sources, we have created a table with the RES support schemes (administratively set FITs or FIPs, TGCs, auctions and tax incentives/investment grants) which have been adopted in the EU member states between 1997 and 2019, with a focus on on-shore wind and solar PV technologies.6 Trends in the implementation of those schemes are shown in Figure 16.2.

30 25 20 15 10

Total

FIT/FIP

TGC

Auction

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

0

1997

5

TI/IG

Notes:   Total = number of countries with at least one support scheme in the respective year; FIT/FIP = administratively set FITs or FIPs (ASFITs/FIPs); TGC = quota with tradable green certificates; auctions; TI/IG = tax incentives/investment grants. The numbers in the “total” category are lower than the sum of the other categories because they refer to the number of countries which have at least one instrument, but some countries combine two of them in given years. Source:   own elaboration based on CEER reports of 2004, 2008, 2011, 2013, 2015, 2016, 2017, 2018 and 2021 and the AURES II database http://aures2project​.eu/

Figure 16.2  Trends in the adoption of support schemes in EU member states between 1997 and 2019

382  Handbook on the economics of renewable energy

We can observe two main patterns and trends. First, a substantial increase in the adoption of deployment support instruments can be observed in the late 1990s/early 2000s, mostly due to their adoption in Eastern European countries. Since then, all countries have supported RES-E in one way or another, although support may not have been available for new installations in specific years. This increase was mostly due to the increased adoption of two instruments: ASFITs/FIPs and quotas with TGCs. Second, regarding the adoption of specific instruments, Figure 16.2 suggests interesting patterns and trends. ASFITs/FIPs have been the dominant instrument in the whole period, except in the last two years (2018 and 2019), in which auctions were the most widespread instrument. In the earlier part of the period, a substantial increase in ASFITs/FIPs could be observed until 2002. This increase was not due to countries changing one instrument for another, but is related to the adoption of an instrument by a country which had not implemented an instrument before. TGCs were virtually non-existent in the first years, but their adoption also increased between 1997 and 2004 to remain constant. Tax incentives/ investment grants have also remained constant, albeit at very low levels, for the whole period. The adoption of auctions experienced an impressive increase between 2015 and 2016, going from only 2 to 15 member states implementing this support scheme, which was most probably the result of the “European Commission Guidelines on State Aid for Environmental Protection and Energy 2014–2020”. This increased adoption of auctions is due to a shift by countries which previously used FITs. Furthermore, the Renewables Directive (2001/2018) states that “Member States shall ensure that support for electricity from renewable sources is granted in an open, transparent, competitive, cost-effective and non-discriminatory manner” (article 4.4), a clear reference to auctions. In the EU, most countries already apply auctions to promote wind energy. To date, according to the AURES II project database (http:// aures2project​.eu/), auctions in which on-shore wind was eligible to participate have been launched in 16 of the 27 member states and concluded in 15 of them (all except Slovakia).7 According to AURES II data, auctions for off-shore wind have been launched and concluded in six countries (Denmark, France, Germany, Italy, the Netherlands and Poland). However, administratively-set FITs still have a significant weight in Europe, as shown in Figure 16.2. Finally, it is important to take into account that some countries combine two schemes (auctions and ASFITs/FIPs, for example), either for different technologies or for the same technologies but for plants of different sizes, with ASFITs/FIPs for smaller projects and auctions for larger ones (as in France). Thus, in the past, most investments in renewable electricity in the EU were promoted under administrative FITs/FIPs schemes, while other primary instruments played a much smaller or even residual role (TGCs and auctions). However, auctions have recently spread across Europe and the rest of the world at very high speed, possibly as a consequence of the advantages they present to public decision-makers in terms of increased control over the evolution of installed capacity and reduced promotion costs by stimulating competition between technologies, projects and developers. This pattern of increased adoption of auctions in the EU is also reflected at the world level. Globally, they have become the flagship instrument: 131 countries were using auctions in 2021 (REN21, 2021), compared to only 6 in 2005 (IRENA, 2019). In comparison, administrativelyset FITs were used in 92 countries in 2021 (REN21, 2021).

The economic analysis of renewable energy policies  383

DESIGN ELEMENTS MAKE A DIFFERENCE: HOW CAN INSTRUMENTS BE DESIGNED AND WHAT ARE THE PROS AND CONS OF DIFFERENT DESIGN ELEMENTS?8 There is a widespread agreement in the literature that the incentive for the diffusion of renewable energy technologies does not only depend on the instrument being used, but also on how the instrument is designed, that is, on its specific design elements. In this section we focus on the design of the primary instruments mentioned above. We do not provide a systematic review of all those design elements per instrument, but only focus on the most relevant ones in order to illustrate their possible influence on the outcomes of the instruments as assessed with the aforementioned criteria. The discussion starts with design elements common to all instruments and follows with an assessment of design elements which are specific to each instrument. Common to All Instruments Technological diversity (neutrality vs. specificity) A main issue in the literature on RES-E support schemes, and a main decision which needs to be taken by policy makers, is whether different renewable energy technologies should be supported or if only the adoption of the currently cheapest should be encouraged. Both have their pros and cons. As mentioned above, technological diversity incentivises technological innovation in a range of technologies, which allows them to reduce their costs in the future, and it also allows different renewable energy resources in a given territory to be captured (especially if it is a large or diverse one). Technological neutrality promises to be cost-effective, i.e., to achieve short-term targets at the lowest possible generation costs (see del Río, 2012; Sandén and Azar, 2005, for a discussion). Under technological neutrality, a similar support level is provided for all technologies (regardless of their generation costs). Under technology-specific support, the remuneration for different technologies is modulated according to those costs. In ASFITs/FIPs, technological neutrality can be implemented through a flat tariff (i.e. the same remuneration level for all technologies). Technological diversity can be promoted by applying technology-specific FITs, i.e., with different support levels per technology according to their generation costs. In TGCs, technological neutrality would be naturally promoted by making all technologies compete without adaptations in the scheme. Technological diversity can be supported by implementing technology-specific quotas and banding. Under a technology-specific quota, targets for different technologies exist, ensuring that less mature, more expensive technologies will be adopted. The assurance of a market for the immature technologies would facilitate the learning effects and RD&D investments in these technologies, encouraging technological innovation, probably at the expense of higher short-term costs. However, if several quotas exist, then narrow market problems may appear, resulting in a too volatile TGC price. A single quota creates an incentive to deploy only the cheapest technologies, i.e., a higher level of technological competition would result. Under banding (also called credit multipliers), more TGCs are awarded to less mature or more expensive technologies. If more TGCs were awarded to immature technologies per MWh of RES-E generation (with the same quota), then a greater effectiveness and lower risk in the deployment of immature technologies (and a greater diversity) would result. There would be more competition under a more technology-neutral scheme

384  Handbook on the economics of renewable energy

and, thus, lower TGC prices. Therefore, as with technology-specific quotas, there is a tradeoff between technological innovation in the longer term and lower short-term generation costs. In auctions, technological neutrality would be implemented by making all technologies compete in the same auction. Technological diversity could be promoted in an auction by organising different auctions per technology or by including minimum quotas per technology (del Río, 2017). Geographical diversity (neutrality vs. diversity of locations) Investors will place their projects where the renewable resource is better. This will lead to a concentration of RES-E projects in specific locations, which reduces private generation costs in terms of a lower LCOE. At first sight, modulating support according to the location of the plant, with greater support levels provided for plants deployed in places with greater costs, may seem at odds with economic efficiency, since installations would not be promoted where generation costs are minimised. Should a diversity of locations be promoted? Diversity would have three main benefits. First, a support level which is not differentiated per location could lead to excessive rents for the owners of plants in places with very good resource conditions. Second, the lower concentration of projects in a given location would increase the social acceptability of RES, by mitigating the NIMBY problem. Finally, diversity may lead to lower indirect costs (grid extensions or reinforcements), if the location of projects near the existing grid is encouraged. Therefore, it may make sense in terms of lower system costs (if dispersion would lead to lower indirect costs, which would more than offset the higher direct costs). In FITs, geographical neutrality would be supported with a flat remuneration level, i.e., the same support for different locations. A stepped FIT would encourage a more even deployment of RES-E projects in the territory. This would provide, for example, a lower level of support in the best places (in terms of renewable energy resource) compared to worse locations.9 A lower level of selection pressure and competition would result. There is not a clear method to support geographical diversity in TGCs, geographical neutrality being the most common alternative adopted. A region-specific TGC scheme could be implemented (as in Belgium), but this segmentation of the market would reduce the efficiency benefits of having a nation-wide TGC scheme (especially if the country is small). Finally, site-specific auctions could be conducted, or project construction could be limited to some previously designated areas. A location-specific correction factor could also be included in the merit order to discourage the construction of locations in specific places (see, e.g., IRENA, 2019). Encouraging diversity could be justified in order to reduce NIMBY effects or to reduce indirect costs, as mentioned above. ASFITs/FIPs ●

●

Capacity or generation caps. Support for ASFITs has generally been provided without a limit on the capacity or generation which could be eligible for support. However, it may be attractive for policy makers to set these limits in order to control the quantity of RES-E projects, as done in quantity-based instruments. This may be so for several reasons: in order to control the costs of support, the amount of electricity that is fed into the grid, etc. In this case, the support would be awarded on a first-come-first-served basis. Tariff degression. Setting the level of support has been a major problem with ASFITs/ FIPs in the past. If set too low, i.e., below the generation costs of the technologies, the scheme would not encourage RES-E investments. If set too high compared to the costs

The economic analysis of renewable energy policies  385

●

●

of generation, excessive remuneration (which is finally paid by consumers) would result. However, the asymmetric information problem makes it difficult for policy makers to set the level of support at appropriate levels, i.e. slightly above the costs of the technology. This problem becomes more serious with high-cost gap or immature technologies which experience very dynamic cost trends in the near future. One way to address this problem is by implementing degression. Fixed degression was first introduced in the German EEG in 2000 and refers to previously set percentage reductions over time in support levels (tariffs) for new plants (degression rates), which involves reductions over time in support levels for new plants. By reducing the financial burden for consumers, degression increases the social acceptability and political feasibility of RES-E support in the long term and, thus, reduces risks. Investors’ risks are reduced if they know a priori how the support level (degression rates) will evolve for the new installations (as in Germany). However, if they are not known ex-ante, they will lead to increased uncertainty for investors (Junginger et al., 2005). Degression encourages innovation in order to reduce the costs of technologies to benefit from the (degressive) tariff. Although it provides an incentive for technological innovation and cost reductions, it cannot accurately correct for sharp declines or extraordinary increases in generation costs. There are two main alternatives to traditional degression: to establish growth corridors, as it was the case with the German FIT, or to link support levels and capacity additions in a circular manner. In these modalities of degression, reductions in support levels partly depend on the capacity installed in the previous year (see del Río, 2012, for further details). Price caps/price floors (only for fixed FIPs). In the case of fixed FIPs, a cap or a floor on the total quantity of support which can be received (price of electricity + premium) can be set. If the price of electricity and the premium are above the cap, the RES-E generator receives the cap. If their sum is below the floor, the generator receives the floor. The cap removes the risks of an excessive remuneration for generators (i.e., high above their costs). The floor mitigates the risks of remuneration that is too low (i.e., lower than their costs). The cap aims to minimise the costs for consumers. The floor aims to encourage the effectiveness of RES-E support. Different FIPs. There are several types of FIPs. A main distinction is between fixed and sliding FIPs. Fixed FIPs are set once and do not alter. The total remuneration thus depends on the market prices. Sliding FIPs are set at regular intervals to fill the gap between the average market price perceived by all generators of a given technology and a reference price, which can be set by policy makers or in an auction. Under a one-sided sliding premium (as implemented in Germany), if the wholesale electricity price is above the reference price, the generators receive the wholesale price. If it is below, the generators receive the reference price. Under a two-sided sliding premium, which is also called a contract-for-differences (as implemented in the UK), if the wholesale electricity price is below the reference price, RES-E generators receive the reference price (see also Chapter 18). If it is above, they have to give back the difference between the wholesale market price and the reference price. Under both fixed and sliding FIPs, generators are required to sell their electricity in the market, which favours the integration of RES-E into the electricity system. However, there is an important difference regarding the risks for investors/generators. These are higher under fixed FIPs, since the total amount of remuneration is previously unknown. This is not the case under sliding FIPs, since generators receive at least the reference price (in the one-sided sliding FIP) or exactly the reference price (in the two-sided sliding FIPs).

386  Handbook on the economics of renewable energy

TGCs ●

●

Maximum and minimum prices. The price of TGCs as a result of the interaction between supply and demand can be either too high or too low compared to the costs of the technologies or the wholesale electricity price. If it is too high, consumers will be overburdened. A low TGC price is beneficial for consumers, but it may have detrimental consequences for the effectiveness of the scheme or its dynamic efficiency, since the less mature technologies would not be promoted (although this problem can be addressed with other design elements, see above). Maximum prices reduce the likelihood that expensive technologies are needed to comply with the quota (mitigating the risks of high costs for consumers), but they also reduce the incentive to invest in immature technologies, resulting in a lower market volume for these technologies. A minimum TGC price would reduce investor risks, although, if activated, it would also result in higher support levels. There would be a higher level of effectiveness but also higher costs for consumers. Banking and borrowing. A major potential problem in TGC schemes is the volatility of its price. Investors need a more or less stable price signal to plan their investments accordingly. The aforementioned maximum and minimum prices can limit such volatility. Another alternative is to use banking and borrowing of TGCs. Banking refers to the possibility of using TGCs issued in one specific year to comply with RES-E targets in a future year. In borrowing, the TGCs which will be issued in a future year can be used to comply with RES-E targets in a previous year. Banking and borrowing are two flexibility mechanisms which prevent price spikes across time (lower volatility) and keep prices more stable and lower than in their absence.10 Banking increases the incentive for (low-cost) RES-E generation in the short term in order to sell the corresponding TGCs at a later date when the price is expected to increase. Borrowing involves a greater supply of TGCs in the short term.

Auctions ●

●

Schedule. Auctions can be conducted on an ad-hoc basis, i.e., without a previous plan on when they will be organised, or with a schedule of approximate auction volumes and dates in the future. As argued in del Río (2017), a long-term schedule for regular auctions published with sufficient anticipation (i.e., three years, depending on the technology) decreases investor risks and encourages participation in the auction. However, ad-hoc auctions provide full flexibility to policy makers to conduct auctions when they believe it is better for their goals. Indeed, as shown by del Río and Kiefer (2021), most countries do not have a schedule in their RES-E auctions. Prequalifications. These are required in order to participate in the bidding procedure and are applied in order to prove the seriousness of bids (Mora et al., 2017). They can refer to specifications of the offered project (such as technical requirements, documentation requirements and preliminary licenses) or to the bidders (providing evidence of the technical or financial capability of the bidding party) (Held et al., 2014). Financial guarantees by participants are often required. As with other elements, the challenge is to set them at appropriate levels (del Río, 2017). If they don’t exist or are too lenient, a lower level of effectiveness of the auction can be expected, since project developers may not build the project if worse conditions occur at the time building takes place (i.e., a lower realisation

The economic analysis of renewable energy policies  387

●

●

rate). If they are too stringent, they lead to higher costs for potential bidders, which results in higher bids (and, thus, higher consumer costs). Participation and, thus, competition would also be affected, leading to the same result: higher bids. See Kreiss et al. (2017) for an in-depth analysis of this topic. Dynamic vs. static auctions. As argued in del Río (2017), auctions can be organised in a static (sealed bid) or a dynamic manner (descending or ascending clock). Dynamic auctions reveal information about prices, which “is particularly relevant when there is uncertainty on the costs of renewable energy projects, mitigating the risk of winners’ curse, which occurs when bidders do not know their actual valuation for the good” (del Río, 2017, p. 9). However, static auctions are less vulnerable to implicit collusion than dynamic ones (Haufe and Ehrhart, 2015). Therefore, not revealing any information during the auction process is an advantage of sealed-bid auctions, especially in the presence of weak competition. Del Río and Kiefer (2021) show that there is a large dominance of static vs. dynamic auctions in the RES-E auctions conducted worldwide. Maximum price limits. Finally, setting ceiling prices may be justified in order to mitigate the risk that awarded bid prices that are too high will result, leading to a high cost for consumers. However, they have to be set at the appropriate level. If set too high, they will not be effective in mitigating the aforementioned risk. If set too low, they will discourage technological diversity (in technology-neutral auctions). An important decision is whether to publish the ceiling prices before the auction or not. As argued by del Río (2017, p. 10), “disclosure would increase transparency, investor confidence, participation and competition (e.g., lower support costs) but it may bias the results of the auction if the bidders propose relatively high bids marginally close to the ceiling price (‘anchoring’)”.

It can be observed that some of the aforementioned design elements lead to a convergence of support schemes in the sense that they put quantity limits on a price-based scheme (e.g., capacity or generation caps) or include price limits in price-based schemes (maximum or minimum prices).

CONCLUSIONS Public policy has been, is and will be a main driver of the decarbonised energy transition. This chapter has provided a structured, critical and historical overview of the evolution of public policy support for renewable electricity technologies, taking into account different stages of the technological change process (development, innovation and deployment), but focusing on the last two decades, deployment support and solar and wind technologies. It has been shown that the economic justification for supporting the development and diffusion of renewable energy technologies has different rationales and that the type of support being provided is contingent upon the situation of the technology in the technology life cycle. It has also been shown that support for renewable energy technologies has several components, beyond the simplistic identification of “policy” with “instruments”. The existence of targets for these technologies and stability of support is as important as the instruments. When it comes to instruments, although the focus in the literature has been on deployment support instruments, RD&D support has played a crucial role in improving the quality and reducing the costs of the currently mature technologies. It has also been shown that, although

388  Handbook on the economics of renewable energy

a wide array of deployment support schemes is available for policy makers to choose from, the implementation of ASFITs/FIPs has been dominant in the past, both globally and in the EU. However, the picture has recently changed, both in the EU and worldwide, and auctions are now the main instrument to support the diffusion of renewable electricity technologies and the deployment of renewable electricity projects. In the near future, deployment support in general and auctions in particular can be expected to play a relevant role in this context, but the maturity and cost-competitiveness of some renewable energy technologies (particularly solar PV and on-shore wind) make it likely that support for them will not be needed in the medium term. Finally, this chapter has illustrated how the success of deployment support instruments strongly depends on the way they are designed, and that trade-offs between different criteria used to assess such success are likely to occur when choosing a given design element over others. Although the qualitative analysis of different design elements has received attention in the literature, future research should be devoted to the quantitative analysis of the impact of those design elements on the functioning of the respective instruments, as well as the tradeoffs involved (see Chapter 18).

NOTES 1. These authors argue that instruments are one element of policies (or policy mixes). They are the concrete tools to achieve overarching objectives and can be seen as tools or techniques of governance that address policy problems (Rogge and Reichardt, 2016, p. 1623). 2. For an analysis of support for research, development and demonstration (RD&D) in the solar PV and concentrated solar power sectors, see Mir-Artigues and del Río (2016) and Mir-Artigues et al. (2019), respectively. 3. There are different types of FIPs, such as fixed FIPs, one-way sliding FIPs and two-way sliding FIPs (also called contract-for-differences) (see the sixth section of this chapter for details). 4. A more detailed description of these criteria and the rationale used to define them is provided in del Río et al. (2015a). 5. The well-known problem of asymmetric information occurs because renewable generators have the information on the costs of the technologies and have a natural incentive to declare higher costs when asked by government authorities with the aim of setting the remuneration levels. 6. This table is not provided here for reasons of space, but it is available from the authors upon request. 7. These countries are Denmark, Estonia, Finland, France, Germany, Greece, Hungry, Ireland, Italy, Lithuania, Malta, the Netherlands, Poland, Slovenia and Spain. In Slovakia a multitechnology auction was scheduled for the 1st of April of 2019 that was cancelled before it took place. 8. This discussion is based on Mir-Artigues and del Río (2016). 9. Nevertheless, when setting the location-specific tariffs, the legislator has to make sure that the profitability of projects at good sites is still higher than at relatively bad sites in order to give producers an incentive to search for the best location. Otherwise, the overall efficiency of the FIT scheme might be seriously lowered (del Río, 2012). 10. Amundsen et al. (2006) show that the introduction of TGC banking may reduce price volatility considerably. Banking lowers average prices but may lead to higher TGC prices in the short-term, however.

REFERENCES Amundsen, E., Bergman, L., & Von Der Fehr, N. 2006. The Nordic electricity markets: Robust by design? In: Sioshansi, F., & Pfaffenberger, W. (eds.), Electricity Market Reform: An International Perspective. Amsterdam: Elsevier, pp. 145–170.

The economic analysis of renewable energy policies  389

Åstrand, K., & Neij, L. 2006. An assessment of governmental wind power programmes in Sweden – Using a systems approach. Energy Policy, 34, 277–296. Cerdá, E., & Del Río, P. 2015. Different interpretations of the cost-effectiveness of renewable electricity support: Some analytical results. Energy, 90, 286–298. Del Río, P. 2008. Ten years of renewable electricity policies in Spain: An analysis of successive feed-in tariff reforms. Energy Policy, 36, 2917–2929. Del Río, P. 2011. Climate change policies and new technologies. In: Cerdá, E., & Labandeira, X. (eds.), Climate Change Policies: Global Challenges and Future Prospects. Cheltenham, UK: Edward Elgar Publishing. Del Río, P. 2012. The dynamic efficiency of feed-in tariffs: The impact of different design elements. Energy Policy, 41, 139–151. Del Río, P. 2017. Designing auctions for renewable electricity support: Best practices from around the world. Energy for Sustainable Development, 41, 1–13. Del Río, P., & Gual, M. 2004. The promotion of green electricity in Europe: Present and future. European Environment, 14, 219–234. Del Río, P., & Kiefer, C. P. 2021. Analysing patterns and trends in auctions for renewable electricity. Energy for Sustainable Development, 62, 195–213. Del Río, P., & Kiefer, C. P. 2022. Which policy instruments promote innovation in renewable electricity technologies? A critical review of the literature with a focus on auctions. Energy Research & Social Science, 89, 102501. Del Río, P., & Linares, P. 2014. Back to the future? Rethinking auctions for renewable electricity support. Renewable and Sustainable Energy Reviews, 35, 42–56. Del Río, P., & Mir-Artigues, P. 2012. Support for solar PV deployment in Spain: Some policy lessons. Renewable and Sustainable Energy Reviews, 16, 5557–5566. Del Río, P., Ragwitz, M., Steinhilber, S., Resch, G., Busch, S., Klessmann, C., De Lovinfosse, I. J. V. N., Fouquet, D., & Johnston, A. 2012a. Assessment criteria for identifying the main alternatives. Deliverable 2.2, Beyond 2020 Project, Funded by the Intelligent Energy-Europe Program. CSIC Madrid. Del Río, P., Ragwitz, M., Steinhilber, S., Resch, G., Busch, S., Klessmann, C., De Lovinfosse, I., Nyszen, J., Fouquet, D., & Johnston, A. 2012b. Key policy approaches for a harmonisation of RES(-E) support in Europe – Main options and design elements. A report compiled within the European research project beyond2020 (work package 2). Supported by Intelligent Energy – Europe, ALTENER (Grant Agreement no. IEE/10/437/SI2.589880). Del Río, P., Haufe, M.-C., Wigan, F., & Steinhilber, S. 2015a. Overview of design elements for RES-E auctions. Report of the EU-Funded AURES Project. Del Río, P., Wigan, F., & Steinhilber, S. 2015b. Assessment criteria for RES-E auctions. Report of the EU-Funded AURES Project. Del Río, P., Kiefer, C. P., Carrillo-Hermosilla, J., & Könnölä, T. 2021. The Circular Economy. Economic, Managerial and Policy Implications. Switzerland: Springer Nature. Edenhofer, O., Hirth, L., Knopf, B., Pahle, M., Schlömer, S., Schmid, E., & Ueckerdt, F. 2013. On the economics of renewable energy sources. Energy Economics, 40, S12–S23. European Commission. 2013. European Commission guidance for the design of renewable support schemes. Accompanying the document Communication from the Commission. Delivering the internal market in electricity and making the most of public intervention. SWD(2013) 439 final. Brussels 5.11.2013. Gowrisankaran, G., Reynolds, S., & Samano, M. 2011. Intermittency and the value of renewable energy. National Bureau of Economic Research. Working Paper 17086. Groenenberg, H., & De Coninck, H. 2008. Effective EU and member state policies for stimulating CCS. International Journal of Greenhouse Gas Control, 2, 653–664. Gross, R., Heptonstall, P., Anderson, D., Green, T., Leach, M., & Skea, J. 2006. The costs and impacts of intermittency: An assessment of the evidence on the costs and impacts of intermittent generation on the British electricity network. A Report of the Technology and Policy Assessment Function of the UK Energy Research Centre. London: UK Energy Research Centre. Harmon, R. R., & Cowan, K. R. 2009. A multiple perspectives view of the market case for green energy. Technological Forecasting and Social Change, 76, 204–213. Haufe, M.-C., & Ehrhart, K.-M. 2015. Assessment of auction formats suitable for RES-E. Report of the EU-Funded AURES Project.

390  Handbook on the economics of renewable energy

Held, A., Ragwitz, M., Gephart, M., De Visser, E., & Klessmann, C. 2014. Design features of support schemes for renewable electricity. A report within the European project “Cooperation between EU MS under the Renewable Energy Directive and interaction with support schemes”. Utrecht, The Netherlands: Ecofys. Holttinen, H. 2011. Design and operation of power systems with large amounts of wind power. Final Summary Report, IEA WIND Task 25, Phase Two 2009–2011. IEA, VTT. IEA. 2021. Renewables 2021. Analysis and forecast to 2026. IRENA. 2019. Renewable Energy Auctions: Status and Trends Beyond Price. Abu Dhabi: International Renewable Energy Agency. IRENA. 2021. Renewable power generation costs in 2020. IRENA. 2022. Renewable power generation costs in 2021. Jacobsson, S., Bergek, A., Finon, D., Lauber, V., Mitchell, C., Toke, D., & Verbruggen, A. 2009. EU renewable energy support policy: Faith or facts? Energy Policy, 37, 2143–2146. Junginger, M., Faaij, A., & Turkenburg, W. C. 2005. Global experience curves for wind farms. Energy Policy, 33, 133–150. Khatib, H., & Difiglio, C. 2016. Economics of nuclear and renewables. Energy Policy, 96, 740–750. Kitzing, L., Fitch-Roy, O., Islam, M., & Mitchell, C. 2018. An evolving risk perspective for policy instrument choice in sustainability transitions. Environmental Innovation and Societal Transitions, 35, 369–382. Kopsakangas-Savolainen, M., & Svento, R. 2013. Economic value approach to intermittent power generation in the nordic power markets. Energy and Environment Research, 3, 139–155. Kreiss, J., Ehrhart, K.-M., & Haufe, M.-C. 2017. Appropriate design of auctions for renewable energy support – Prequalifications and penalties. Energy Policy, 101, 512–520. Markard, J., & Truffer, B. 2008. Technological innovation systems and the multi-level perspective: Towards an integrated framework. Research Policy, 37, 596–615. Mendonça, M., Lacey, S., & Hvelplund, F. 2009. Stability, participation and transparency in renewable energy policy: Lessons from Denmark and the United States. Policy and Society, 27, 379–398. Mir-Artigues, P., & Del Río, P. 2016. The Economics and Policy of Solar Photovoltaic Generation. Switzerland: Springer. Mir-Artigues, P., Del Río, P., & Caldés Gómez, N. 2019. The Economics and Policy of Concentrating Solar Power Generation. Switzerland: Springer. Mitchell, C., Sawin, J. L., Pokharel, G. R., Kammen, D., Wang, Z., Fifita, S., Jaccard, M., Langniss, O., Lucas, H., Nadai, A., Blanco, R. T., Usher, E., Verbruggen, A., Wustenhagen, R., & Yamaguchi, K. 2011. Policy, financing and implementation. In: Edenhofer, O., Pichs-Madruga, R., Sokona, Y., Seyboth, K., Matschoss, P., Kadner, S., Zwickel, T., Eickemeier, P., Hansen, G., Schlomer, S., & Von Stechow, C. (eds), IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation. Cambridge: Cambridge University Press. Mora, D., Kitzing, L., Rosenlund, E., Steinhilber, S., Del Río, P., Wigand, F., Klessmann, C., Tiedemann, S., Amazo, A., Welisch, M., Kreiß, J., Fitch Roy, O., & Woodman, B. 2017. Auctions for renewable energy support – Taming the beast of competitive bidding. Final Report of the EU-Funded AURES Project. Ortega-Izquierdo, M., & Del Río, P. 2020. An analysis of the socioeconomic and environmental benefits of wind energy deployment in Europe. Renewable Energy, 160, 1067–1080. Ragwitz, M., Steinhilber, S., Breitschopf, B., Resch, G., Panzer, C., Ortner, A., & Busch, S. 2012. RE-shaping: Shaping an effective and efficient European renewable energy market. Final Project Report. RE.Shaping Project. Karlsruhe: Fraunhofer ISI. Rathmann, M. 2011. Towards triple – A policies: More renewable energy at lower cost. A report compiled within the European research project RE-Shaping (work package 7). Intelligent Energy – Europe. Reichardt, K., Rogge, K. S., & Negro, S. O. 2017. Unpacking policy processes for addressing systemic problems in technological innovation systems: The case of offshore wind in Germany. Renewable and Sustainable Energy Reviews, 80, 1217–1226. REN21. 2021. Renewables Global Status Report. Paris: REN21 Secretariat. Rogge, K. S., & Reichardt, K. 2016. Policy mixes for sustainability transitions: An extended concept and framework for analysis. Research Policy, 45, 1620–1635.

The economic analysis of renewable energy policies  391

Sandén, B. A., & Azar, C. 2005. Near-term technology policies for long-term climate targets - Economy wide versus technology specific approaches. Energy Policy, 33, 1557–1576. Stern, N. 2007. The Economics of Climate Change: The Stern Review. Cambridge: Cambridge University Press. Stram, B. N. 2016. Key challenges to expanding renewable energy. Energy Policy, 96, 728–734. Verbruggen, A., & Lauber, V. 2009. Basic concepts for designing renewable electricity support aiming at a full-scale transition by 2050. Energy Policy, 37, 5732–5743.

17. Renewable energy auctions: an overview Vasilios Anatolitis and Jenny Winkler

INTRODUCTION By 2021, more than 130 countries had used auctions or tenders1 to allocate and determine the levels of support for electricity from renewable energy (RE) sources, making them one of the predominant RE support instruments worldwide (REN21 2022). Typically, RE auctions have been implemented due to their presumed advantages for policymakers: the efficient control of RE deployment volumes, reduced support expenditures and the increased market integration of renewables (del Río and Linares 2014; Fitch-Roy et al. 2019; IRENA 2015). On the other hand, opponents of RE auctions stress the drawbacks of using auctions in the RE sector: among others, reduced actor diversity (especially with regard to smaller actors) (IRENA 2015), the potential non-achievement of RE targets and an increased risk for project developers (Đukan and Kitzing 2021). There is still an ongoing debate about whether RE auctions have been able to fulfil their promises (del Río and Kiefer 2023). Therefore, given the importance of auctions in the RE landscape and based on the existing literature on the topic (see del Río and Kiefer 2023), this chapter provides an overview of RE auctions. The first subchapter presents the history of RE auctions. The second subchapter describes the concept of RE auctions and discusses how they work. The third subchapter provides an overview of the relation between RE auctions and several policy objectives. The fourth subchapter presents the possible auction design elements and their impact on several assessment criteria. The fifth subchapter discusses the emergence and potential approaches to zero-price bids in RE auctions. Lastly, the sixth subchapter provides an outlook on the future of auctions.

A BRIEF HISTORY OF RENEWABLE ENERGY AUCTIONS The first countries in Europe to have introduced auctions to support RE were the UK, Ireland and France in the 1990s. Between 1990 and 1998, the UK conducted five auctions under the “Non-Fossil Fuel Obligation” (NFFO) (Mitchell 2000), while Ireland held several auctions under the “Alternative Energy Requirement” (AER) between 1995 and 2003 (Steinhilber 2016a). France conducted several RE auctions under the “EOLE 2005” programme between 1996 and 2007 (del Río and Linares 2014). While the three schemes achieved relatively low prices for RE, the auctions suffered from a low realisation rate of RE projects. Due to this disappointing outcome, Ireland and the UK switched to other RE support instruments, while other European countries abstained from using auctions (del Río and Linares 2014). Outside of Europe, RE auctions also have a long tradition (see, e.g., del Río and Kiefer 2021). After the UK’s NFFO, China implemented the wind power concession policy in the years 2003–2007. They were followed by the Brazilian RE auctions that were introduced in 2007 and are still in place (Bayer 2018; Förster and Amazo 2016; Diniz et al. 2023). Further 392

Renewable energy auctions 

393

notable examples can be found in Latin America: Uruguay, Argentina and Peru already conducted their first RE auctions in 2009, with many more countries following suit. In Europe, after the rather negative experience of Ireland, France and the UK, the main discussion in the realm of RE support instruments was between price- and quantity-based support instruments. The former consisted mainly of the administratively set feed-in tariffs, while the latter consisted mainly of quota schemes with tradable green certificates (TGC) (see Chapter 16). After intensive discussions between proponents in both academia and politics, the administratively set FITs were the “winners”, mainly due to their success in deploying larger shares of volumes at lower costs (Haas et al. 2011). Some technical issues with renewables (negative electricity market prices, forecasting errors, etc.) led many countries and the European Commission to start introducing feed-in premiums (FIPs) in order to incentivise higher market integration for RE generators by exposing them to price signals. After the financial turmoil at the beginning of the 2010s, both in a general sense (financial crisis, etc.) and with regard to the RE sector (deficits in the RE accounts and suspension of RE support schemes, e.g., in Spain or Greece), countries strived for more financially stable RE support schemes. Since they aimed to have more control of the deployment volumes of RE, while at the same time reducing the required support expenditures, introducing auctions was almost a natural consequence. Additionally, RE auctions were in line with the preferences of parts of the EU Commission, which favoured competitive support policies (Fitch-Roy et  al. 2019). Although certain countries, such as France (in 2005 and 2011) (Förster 2016; Winkler et al. 2018), Portugal (in 2006) (del Río 2016) and the Netherlands (in 2011) (Jakob et  al. 2019; Noothout and Winkel 2016), started introducing RE auctions, the major driver for RE auctions in Europe was the EU Commission’s State Aid Guidelines in 2014 (European Commission 2014). The State Aid Guidelines required EU member states to introduce auctions to allocate support for large-scale RE projects, starting in 2017. Therefore, by 2021, 19 EU member states had adopted RE auctions as their main support instrument (AURES II 2022). In North America, RE auctions have been used as an RE support instrument on the regional level. For instance, California used auctions to help suppliers procure RE capacity with a standardised auction procedure (Fitch-Roy 2015). In Canada, mainly Alberta has used auctions to support RE deployment (Menzies and Marquardt 2019; Hastings-Simon et al. 2022). In contrast, Mexico has conducted three RE auction rounds between 2015 and 2017 to procure RE capacities and to move gradually towards a liberalised electricity system (del Río 2017a, 2019b). Nevertheless, in 2018, the Mexican government decided not to conduct any new auction rounds (IRENA 2019). In Australia, RE auctions are usually implemented at the regional level, for instance in the Australian Capital Territory (ACT), in Queensland or in Victoria (Greg et  al. 2019; IRENA 2019). Africa is also gaining ground in this context, with several countries having already conducted RE auctions. The first two countries to experiment with RE auctions were Morocco, which started auctioning RE projects in 2010, and South Africa, which introduced its “REIPPP” scheme in 2011 (del Río and Linares 2014). In recent years, Sub-Saharan countries have started introducing RE auctions as well, such as Uganda and Zambia (Kruger and Eberhard 2018). In Asia, India and China are the countries with the largest RE capacity additions through RE auctions (IRENA 2019). China has been one of the first-comers of RE auctions, with the first auction held in 2003 (Steinhilber 2016b). Lastly, countries in the Middle East have started introducing RE auctions, as well, with the UAE being the frontrunners, having already

394  Handbook on the economics of renewable energy

auctioned CSP projects in 2010 (del Río et al. 2019a). Starting in the late 2010s, other countries in the Gulf Region, such as Qatar and Saudi Arabia, have also adopted RE auctions, resulting in record-low prices (Al-Sarihi and Mansouri 2022). To sum up, auctions have become an important instrument for RE support in each region of the world, with an increasing number of countries having plans to introduce RE auctions in the near future, such as Austria (Resch et al. 2022).

WHAT ARE RENEWABLE ENERGY AUCTIONS? An auction is a competitive mechanism that allocates and prices one or several goods. Auctions can be distinguished between “seller” auctions, where one seller offers a good to several buyers, and “procurement” auctions, in which one buyer procures a good from several sellers. Renewable energy auctions, or more specifically, auctions for the support of electricity from RE, are a market-based mechanism to allocate support to RE projects. Typically, the government auctions a certain volume and interested project developers propose a specific amount of support, which they require to realise their intended RE project. Thus, RE auctions can be classified as “reverse” or “procurement” auctions, as the auctioneer, typically the government, is not selling but procuring a specific good. In these reverse auctions, the auctioneer aims to achieve the lowest possible price. In practice, RE auctions are usually combined with (and are used to set) the support levels of a generation-based remuneration scheme, i.e., a FIT or a FIP (Kitzing et al. 2012). Only in a few cases have RE auctions been combined with investment-based remuneration schemes (e.g., in the Spanish auctions in 2016 and 2017). On the one hand, auctions used in the context of renewable energy support offer two main advantages compared to administratively set support schemes: (1) they allow governments to control the supported RE capacity efficiently and (2) they set the support levels competitively (del Río and Linares 2014). Under administratively set support schemes, governments try to estimate the costs of RE projects and offer this amount as support to RE projects. If this level of financial support is sufficient, project developers will go forward and realise their projects. Nevertheless, the government has only ex-post information on whether the support levels were adequate: if these were set too low, there might be no interest from project developers, while if they are set too high, overcompensation might occur and too much RE capacity might be realised. While the latter might be perceived as positive in terms of RE target achievement, it might lead to excessive support costs, which can increase levies for households or lead to an increase in tax rates. Thus, auctions are especially suitable if information asymmetry between two actors exists, such as between the government and the RE project developers, on the cost and the amount of support needed for electricity generation from RE. Consequently, the main advantage of auctions lies in their theoretical and perceived ability to allocate a certain good efficiently, which is one of the main reasons for their widespread adoption (Fitch-Roy et al. 2019; Grashof 2021). Furthermore, governments can hardly control the supported volumes in administratively set schemes in an economically efficient manner: while governments could set an overall capacity or budget limit for the support, projects with the highest generation costs may apply for the support (earlier), leaving the less expensive projects without support and thus unrealised. At the same time, there are caveats for both bidders and the auctioning entities. For bidders, participating in an auction is complicated, requires specific knowledge and the award is unknown, which is why auctions can discourage smaller players such as households or small

Renewable energy auctions 

395

project developers from participating (Côté et al. 2022; Grashof 2019). The risk of not winning the auction can also imply higher costs due to risk premiums (Đukan and Kitzing 2021). The auctioning entity might face the challenge that without sufficient competition in the auction, bidders can submit relatively high bids, harming support cost efficiency (Batz Liñeiro and Müsgens 2023). Also, auctions might harm the target achievement if not all awarded projects are built, such as in the case of the UK in the 1990s (Fitch-Roy et al. 2019; Mitchell 1995, 2000). Thus, while RE auctions can in some instances reduce support expenditures and help achieve the RE targets, they are not necessarily always superior to other support instruments (Winkler et al. 2018).

RENEWABLE ENERGY AUCTIONS AND POLICY OBJECTIVES Similar to other policy instruments, policymakers usually aim to achieve certain objectives with their RE auctions. In the case of RE auctions, these can usually be divided into primary and secondary objectives (see also Chapter 16). Primary objectives usually include efficiency and effectiveness. While the term efficiency is used extensively in the RE support literature, it appears that different meanings and definitions are attributed to this expression. In the following, drawing from del Río and Cerdá (2014) and other relevant studies, we provide an overview of these different interpretations. First, the literature typically differentiates between a short- and long-term focus of efficiency. The latter is usually referred to as dynamic efficiency. This term takes into account the long-term cost effects of the technologies, including innovation and technology cost reductions (e.g., del Río 2017b), and entails the idea that the received support stimulates the technology learning curve and reduces the cost of technologies that are at the moment rather expensive (del Río et al. 2015). The literature uses the term static efficiency when focusing on the short-term perspective. In general, static efficiency is understood as “reaching a target at minimum costs” (del Río and Cerdá 2014). Depending on the perspective and the definition of costs, two concepts are referred to as efficiency in the RE support literature: the minimisation of RE electricity generation costs and the minimisation of support costs. The first concept is typically defined by RE support literature as achieving a specific RE target at the lowest electricity generation costs. In practice, this means that efficiency is reached (in accordance with the equi-marginality principle) if RE projects with the lowest generation costs are deployed (Ragwitz and Steinhilber 2014). From a welfare economics perspective, this means that the overall social welfare is maximised, making this outcome desirable from the society’s point of view. Standard economic literature typically refers to this concept of efficiency as Pareto efficiency or, in a similar vein, allocative efficiency. Several studies from the theoretical literature on RE auctions (e.g., Kreiss et al. 2017a, 2017b) follow the approach of allocative efficiency. Usually, this concept is simply referred to as static efficiency, costeffectiveness, or cost efficiency. In the context of RE auctions, we understand that an efficient outcome in this sense is achieved if the projects with the lowest levelised cost of electricity (LCOE) are awarded in the auctions (Haufe and Ehrhart 2018). Consequently, in this chapter, we will refer to static efficiency, if projects with the lowest LCOE are awarded in the auctions. The second concept found in RE support literature interprets efficiency as minimising the support expenditures of the government when achieving a specific RE target. In welfare economics, support expenditures are merely transfers between producers and consumers and are typically disregarded. Minimising the support expenditures simply means that

396  Handbook on the economics of renewable energy

consumers’ surplus is increased at the expense of producers’ surplus, while the overall welfare is unchanged. Nevertheless, these distributional effects are deemed important in the RE support literature, as these impact the burden of electricity consumers, raising challenges, e.g., concerning socially vulnerable households or the competitiveness of energy-intensive industries. Early RE auction literature sometimes included this objective in the term static efficiency or referred to the concept simply as “minimisation of support costs” (del Río et al. 2015). However, recent studies in the RE auction literature have adopted the term support cost efficiency (e.g., Álvarez and del Río 2022; Anatolitis and Welisch 2017; Tolmasquim et al. 2021), which is defined as the objective of achieving low awarded prices in the auctions (e.g., Álvarez and del Río 2022; Anatolitis et al. 2022; Ehrhart et al. 2019) or, more precisely, awarding projects at the lowest prices possible.2 We consider support cost efficiency as the more relevant and more researched objective in the context of RE auctions. First, for (empirical) studies on RE auctions, it is difficult to observe the actual generation costs of the participating projects. Second, in recent years, reducing support costs has become a more decisive policy objective for policymakers than the rather abstract and theoretical objective of reducing generation costs/achieving static efficiency. In addition to (support cost) efficiency, effectiveness plays a crucial role in assessing RE auctions. Effectiveness is usually understood as the ability of a support instrument to achieve a specific target. In the RE support literature, effectiveness often “refers to the extent to which a promotion strategy is capable of triggering RE deployment, either measured in increased generation or increased installed capacity” (Ragwitz and Steinhilber 2014, p. 220). A more narrow definition of effectiveness can be found in the RE auction literature, which is particularly relevant for the case of auctions: most studies refer to it as achieving a high realisation rate of awarded projects, sometimes referred to as ex-post effectiveness (del Río 2017c, 2018). Further, the concept of a priori effectiveness is also used, which refers to the ability of the auctioneer to award the offered capacity/generation, meaning the auction is not undersubscribed (del Río 2017c, 2018). Given the negative experiences of low realisation rates of awarded projects in the past, the RE auction literature acknowledged the importance for policymakers of safeguarding effectiveness when striving for other policy objectives (Kitzing et al. 2019). Secondary objectives are manifold but usually include actor diversity, economic growth or security of supply. The policy objective of actor diversity refers to the objective of guaranteeing a diverse field of players (del Río 2017b). In the RE auction literature, this typically means awarding projects from small-scale players, e.g., energy communities. Another secondary objective can be industrial development through an increasing share of RE, sometimes referred to as green growth. Typically, green growth is understood as (shortterm) economic growth, especially increased employment induced by investments in renewable energy technologies (Eicke and Weko 2022; Huberty et al. 2011). Ensuring the reliable supply of electricity can be identified as a policy objective of RE support policies. This security of supply is usually defined as providing a stable electricity system (sufficient electricity supply) and power grid in the short-term (Paravantis and Kontoulis 2020; r2b energy consulting GmbH 2019), e.g., to avoid outages. Further, security of supply can be achieved if RE installations are spread across different regions and are not concentrated in a specific area. Thus, security of supply is closely linked to the objective of geographical diversity, which aims at RE projects being installed over several locations and regions. In addition to those mentioned above, several other policy objectives exist, such as system cost efficiency or social acceptability (see Chapter 16).

Renewable energy auctions 

397

Most of these policy objectives can interact, leading to synergies or even conflicts between each other (del Río et al. 2015). Howlett (2009, p. 74) refers to these synergetic policy objectives as coherent: “successful policy design requires (1) that policy aims, objectives, and targets be coherent”. In case of conflicting (and so incoherent) policy objectives, such as support cost efficiency and green growth, policymakers typically prioritise one certain policy objective when designing their support instruments (Fleck and Anatolitis 2023).

RENEWABLE ENERGY AUCTION DESIGN Auctions seem to be a suitable instrument if the minimisation of support expenditures (or generation costs) is the main priority. In contrast, administratively set FITs/FIPs are typically more suitable when effectiveness or actor diversity is the main focus (Held et al. 2014) or in the early market deployment phase (Kitzing et al. 2020). At the same time, not only the support instrument choice but also its design has an impact on the ability to pursue a policy objective. As Ragwitz and Steinhilber (2014, p. 217) stated, “the success of a policy depends not only on the instrument chosen but to a large extent on its concrete design elements”. The high number of possible design elements of RE auctions makes the design process quite challenging, especially since the auction design elements can have contrasting effects on the assessment criteria (Fleck and Anatolitis 2023). Therefore, the following subchapters present (1) some guiding principles for auction design, followed by (2) an extensive list of RE auction design elements and their effect on the various policy objectives. Guiding Principles for Designing RE Auctions The design of RE auctions should reflect the desired policy objectives and external circumstances. Policymakers typically pursue specific objectives when drafting their national RE legislation. Thus, the effects of the chosen auction design elements should favour these objectives. An auction design that reflects the pursued objectives can avoid disappointing auction outcomes for policymakers. Having coherent policy objectives in place, meaning objectives that can easily be achieved simultaneously, such as effectiveness and green growth, makes the process of designing auctions less challenging. Otherwise, incompatible policy objectives might need to be prioritised, and policymakers need to navigate the various trade-offs when designing auctions. Furthermore, a no “one size fits all” approach for auction design exists, as the auction design is highly contextual and should reflect the national market environment in each country (Kitzing et al. 2019; Mora et al. 2017a, 2017b). The auction needs sufficient competition to function properly. Auctions as an allocation mechanism only work correctly under enough competition. In the context of RE auctions, this means that a high number of projects need to participate in the auctions. However, it should be noted that it is up for debate which specific level of competition can be considered sufficient. Nevertheless, policymakers should avoid implementing auction design elements that artificially create competition, e.g., the automatic volume adjustment, due to potential adverse effects in the mid- to long term (Hanke and Tiedemann 2020). The automatic volume adjustment reduces the original auctioned volume in an undersubscribed auction, so the submitted bids’ total capacity surpasses the final auctioned volume (for further details, please see later in this chapter). Therefore, it is ensured that there will be always some bidders that will not be awarded. This removes the incentive for participating in the auctions for bidders with high

398  Handbook on the economics of renewable energy

generation costs, thus reducing the overall participation in the mid- to long term (Ehrhart et al. 2020). Policymakers should instead focus on reducing administrative and regulatory barriers for RE projects, e.g., streamline permitting procedures or increase site availability. The risk of the “winner’s curse” should be reduced. The “winner’s curse” occurs if bidders underestimate their costs and/or overestimate future (electricity market) revenues and bid very low prices in the auctions. Thus, bidders with the most optimistic expectations are more likely to be awarded in the auction. Nevertheless, if the actual developments do not meet these expectations, awarded bidders might not realise their project, as they would incur a loss. Auction design should therefore try to reduce the likelihood of the winner’s curse occurring. The auction design should be kept as simple as possible. Policymakers have many degrees of freedom and a large set of potential auction design elements to include when designing their RE auctions. Nevertheless, the experience with RE auctions has shown that the simpler an auction is designed, the lower the transaction costs for participating. Thus, a simple auction tends to attract more project developers and investors, leading to a higher level of competition and, thus, lower awarded prices. Impact of Auction Design Elements on Policy Objectives Table 17.1 provides a comprehensive overview of selected auction design elements for RE auctions and scores its effects on the policy objectives. A brief introduction and discussion of the effects of each design element are provided in the rest of this section. Our analysis draws from and extends the work of del Río (2017b), del Río and Kiefer (2021) and Haelg (2020), who have already provided comprehensive overviews of RE auction design elements. The large variety of design elements gives policymakers several degrees of freedom. However, it also poses the challenge of navigating through their trade-offs. For instance, a floor price, which defines a lower level for the submitted prices, is suitable for effectiveness, as bidders cannot submit unsustainably low prices. At the same time, a floor price obviously harms support cost efficiency, since submitted bid prices could be potentially lower than the floor price. Due to these different effects, design elements should be carefully chosen and typically reflect the stated policy objectives. Typically, not only the choice of the design element has an impact on the policy objectives, but its actual implementation. One such example is the reference market value under a onesided sliding FIP (for a description of different types of FIPs, see below and also Chapter 16). Most countries in the EU have implemented a sliding FIP, which usually needs a proxy for the RE generators’ electricity market income, often referred to as the reference market value. Thus, when applicable, we also present the different implementation options. Auction scope Auctioned volume The auctioned volume describes the overall volume up for auction and typically represents the volume of RE projects the government intends to support. In the case of multi-item auctions, the auctioned volume can be defined as capacity (in MW), electricity (in MWh) or budget (in EUR/USD/other currencies) (del Río and Kiefer 2021). The higher the auctioned volume, the more projects are awarded, which leads (1) to more expensive projects being awarded and (2) to a lower level of competition. Both aspects lead to higher awarded bid prices and, thus, a lower support cost efficiency. The same accounts for the static efficiency. In contrast, awarding a higher volume might award more expensive projects/

399

+ – +(/–) +

↑

↓

↑

Price-only

Floor price

Ceiling price

Automatic volume adjustment

Selection criteria

+

–

Uniform

+ –(/+)

Static

Pay-as-bid

Pricing rule

Auction type

Dynamic

–

↑

Local content requirements

Allocation process

– –

↑

↑

Financial prequalification

–

↓

Material prequalification

+(/–)

↑

–(/+)

↑

Maximum project size

+(/–)

Multi-technology

Minimum project size

Qualification requirements

Location specificity

–

–

Multi-item

Technology-specific

Single-item

Technological diversity

+

↑ –

Support cost efficiency

Auction format

Design option

Auctioned volume

Auction scope

Design element

Table 17.1  Overview of selected auction design elements

+

–

–

–

–

–

–

+

–

+

–

–

Static efficiency

–

–

–

+

+

Dynamic efficiency

–

+

–

+

+

–

–

+

+

+

–

+

Effectiveness

–

–

–

+

+

–

+

–

–

–

+

–

–

+

+

–

+

Actor diversity

–

+

+

–

+

Green growth

(Continued)

–

+

–

+

Security of supply

400

+ +/– +/– +(/–) + + –

Contract-for-difference

Fixed FIP

Investment grant

↑

↑

↑

↑

Support duration

Flexibility

Realisation period

Penalties

–

–

Static efficiency +

(+)

Dynamic efficiency

+

–

–

–

–

+

+/–

+

+

–

+

+

Effectiveness

–

+

+

–

–

+

+/–

+

+

+(/–)

Actor diversity (+)

Green growth (+)

Security of supply

Note:   (+) means positive effect of the design element on the respective criterion; (–) means negative effect of the design element on the respective criterion. (↑) means that the value of the design element increases, while (↓) indicates that its value decreases.

+ +/–

Feed-in tariff (FIT)

–

Next highest

One-sided sliding FIP

Remuneration scheme

Support and contract design

+ –

– –

↑

Entirely

Bonus/malus

Last awarded bid

Up to the auctioned volume

–

Multi-criteria

Not awarded

Support cost efficiency

Design option

Design element

Table 17.1  (Continued)

Renewable energy auctions 

401

technologies that might become cheaper in the future, which is favourable for dynamic efficiency. Nevertheless, lower auctioned volumes can increase the competition and lead to more aggressive bidding, which might lead to the “winner’s curse” for the awarded bidders. This can harm realisation rates and, thus, effectiveness. Higher auctioned volumes increase the chance that smaller bidders are awarded, thus encouraging actor diversity. Auction format Regarding the auction format a distinction between single- and multi-item auctions can be made (Haufe and Ehrhart 2018). In single-item auctions, only an individual project to be developed is auctioned. For instance, the government can auction a specific site for the development of an offshore wind project. Typically, only one bidder is awarded in the auction. In multi-item auctions, several bidders participate with their individual projects and compete for support. Typically, several different bidders are awarded. In single-item auctions, when only one large project is auctioned, bidders might profit from economies of scale, which can lead to lower bid prices. Nevertheless, if only one bidder is awarded, underbidding might occur which might endanger the realisation of the project and thus harm effectiveness. Awarding multiple bidders/projects mitigates the chance of nonrealisation. Additionally, having multiple awarded bidders increases the chances that smaller project developers are awarded, which increases actor diversity. Technological diversity Policymakers can determine how many technologies compete against each other for support (del Río and Kiefer 2021; Kreiss et al. 2021). Design options include technology-specific auctions, in which only one technology participates, and multi-technology auctions, in which several technologies compete against each other. Multi-technology auctions can be further differentiated into technology-basket auctions, in which similar technologies compete for support, for instance, onshore wind and PV, or in technology-neutral auctions, in which theoretically all RE technologies can participate without any discrimination. Another design option for policymakers is to introduce quotas for certain technologies in multi-technology auctions to ensure that a minimum/maximum volume of a certain technology is awarded. In theory, the more technologies compete against each other, the better the static and support cost efficiency, as the most competitive projects are awarded. Nevertheless, the support cost efficiency (and to a lesser extent, the static efficiency) can suffer, in case of a significant difference in generation costs between the two technologies: projects of the cheaper technology can increase their bid prices to the level of the more expensive technology, thus realising windfall profits. Introducing technology-specific ceiling prices could mitigate this adverse effect. Regarding dynamic efficiency, technology-specific auctions are advantageous, as technologies that might become cheaper in the future might not be competitive in the short term in multi-technology auctions. Furthermore, technology-specific auctions have a rather positive effect on actor diversity, as smaller project developers have challenges in dealing with intertechnology competition. Additionally, technology-specific auctions can positively impact green growth, if policymakers focus on auctioning the technology with the highest potential of domestic economic growth, as well as on security of supply, as technology-specific auctions ensure a variety of technologies will be awarded and most probably deployed (Melliger 2023). Location specificity Policymakers can decide to restrict supported projects to specific locations (Haelg 2020). In the strictest case, policymakers can conduct single-item auctions for a specific site, as, for

402  Handbook on the economics of renewable energy

example, in the Danish or German offshore wind auctions. In most cases, the site is already pre-developed by the government. Alternatively, policymakers can restrict the projects’ locations to specific regions/locations (e.g., solar parks in the Indian PV auctions (IRENA 2019; Probst et al. 2020)) or to certain grid access points (e.g., in the Portuguese RE auctions (del Río et al. 2019b)). Similar to technological diversity, policymakers can introduce quotas to safeguard/restrict the number of projects from a certain region. Restricting the project developers’ freedom to choose the optimal site for their project can lead to higher generation costs and thus higher bid prices, which harms the support cost efficiency. Nevertheless, if the government chooses a site with optimal resources, reduces the transaction costs through central pre-development and reduces regulatory risks through guaranteed grid connections, site- or location-specific auctions can lead to lower awarded prices, improving support cost efficiency. Nevertheless, static efficiency might suffer if the auction is restricted to specific sites, as project developers could have found more optimal sites with lower generation costs. In terms of security of supply, it might be advantageous if the government controls the deployment of RE projects in areas with already congested grids. Qualification requirements Minimum project size The auctioneer can introduce both minimum and maximum thresholds for the project sizes that can participate in the auction (Haelg 2020). These can be set in terms of the project’s capacity or generated electricity. Projects below the minimum size restrictions are not considered in the auction process. Minimum size restrictions exclude smaller projects that have higher generation costs and thus higher bid prices. Thus, a minimum size can have a positive impact on support cost efficiency, as well as static efficiency. Moreover, too strict minimum size restrictions can exclude a large number of smaller projects, which can harm competition, and thus static efficiency, and in addition, harms actor diversity. Maximum project size Similarly to minimum project sizes, the auctioneer can introduce maximum size restrictions (Haelg 2020). Again, these can be set in terms of the project’s capacity or generated electricity. Projects above the maximum size restrictions are not considered in the auction process. Introducing a (too strict) maximum size can impede large-scale projects, which usually profit from economies of scale. Thus, maximum size restrictions tend to harm support cost efficiency. Similarly, maximum size restrictions can harm static efficiency. Nevertheless, maximum size restrictions can lead to a more diverse field of project developers being awarded, and thus, to increased actor diversity. Furthermore, maximum size restrictions can lead to higher green growth, as smaller projects tend to be more labour intensive and thus create more jobs. Financial prequalification Financial prequalifications are financial securities, mostly in the form of bid bonds or completion bonds, that ensure the seriousness of the bids and are often accompanied by penalties for non-realisation of an awarded project, which they should cover (Kreiss et al. 2017b). Typically, if the bid is not successful in the auction and if the awarded project is realised, bidders receive their bond back. As bidders have to provide financial securities to fulfil the requirements, participation in the auction is costly for bidders, and the associated risk increases, leading to higher bid prices,

Renewable energy auctions 

403

harming support cost and static efficiency. Nevertheless, financial prequalifications reduce the incentives for underbidding and thus increase the effectiveness. Smaller players might not be able to raise enough money to submit the bid bonds and are therefore excluded from the auction, which harms actor diversity. Material prequalification Material prequalification requirements should ensure that bidders are capable of realising their project and that the project is already in an advanced stage of development (Kreiss et al. 2017b). Required documents usually consist of building permits, grid connection agreements or environmental impact assessments. Similar to the financial prequalifications, bidders have to fulfil specific requirements to participate in the auctions, which is costly and thus increases the generation costs and the bid prices, harming static and support cost efficiency. Nevertheless, material prequalifications increase the probability of the project being developed, as they shield the project developer from unexpected events; thus, effectiveness is supported. Smaller players might face challenges meeting the material prequalification requirements (before being awarded), which is why actor diversity might suffer. Local content requirements Local content requirements mandate that certain components of the RE projects are sourced from domestic producers, e.g., in the Indian RE auctions (Probst et al. 2020), or that a specific share of the overall investment volume is spent locally, e.g., in the auctions in Saudi Arabia. Different design options for local content requirements exist, each with a certain degree of strictness. For instance, the auctioneer can impose that project developers must spend a certain share of the overall investment volume locally (Hansen et al. 2020). A stricter approach prescribes that project developers need to source certain components of their projects locally, such as the PV modules or wafers. Both design options can be designed more or less strictly, meaning that they can be mandatory, i.e., bidders cannot participate in the auction or lose the award if they do not fulfil the requirements, or the local content requirements can be voluntary, meaning that bidders receive a certain bonus if they meet the requirements, making the non-fulfilment less significant. If the domestic supply chain is still developing, i.e., local producers cannot meet the demand, and there is low competition between suppliers, the CAPEX of RE projects tend to be high, which leads to higher bid prices. In contrast, if the domestic supply chain is already welldeveloped, the effect of local content requirements on the investment costs can be marginal. Nevertheless, the use of local content requirements, in this case, might be arbitrary. Thus, local content requirements generally tend to increase the awarded prices in RE auctions, harming support cost efficiency. Furthermore, local content requirements might affect static efficiency negatively as the projects’ generation costs increase. Nevertheless, one can argue that by increasing the overall investment costs for project developers, the overall generation costs are increased, and thus static efficiency is decreased. Local content requirements can cause delays and nonrealisation of awarded projects, if the domestic market is not able to deliver the needed components. Moreover, local content requirements tend to have a negative impact on actor diversity, as smaller actors, as well as foreign project developers, might be discouraged from participating in the auctions as they might have challenges meeting the foreseen criteria. Finally, local content requirements have a positive impact on domestic manufacturing capacities, leading to increased economic activity in the local RE supply chain, at least in the short term.

404  Handbook on the economics of renewable energy

Allocation process Auction type The auction process can either be static or dynamic (Haufe and Ehrhart 2018). In static auctions, bidders typically submit one bid price and cannot change it afterwards. In contrast, in dynamic auction formats, bidders adapt their bid prices continuously, either until the auctioned volume is reached or a certain deadline has passed. Dynamic auctions might encourage bidders to collude, while static auctions can lead to lower awarded bid prices and thus higher support cost efficiency. Nevertheless, dynamic auctions might be favourable for support cost efficiency, as well, as they enable real price discovery. In contrast to static auctions, the price discovery in dynamic auctions can reduce the risk of underbidding, which supports the effectiveness. Nevertheless, dynamic auctions are more complex and thus induce higher transaction costs, potentially discouraging smaller players from participating and thus harming actor diversity. Pricing rule The pricing rule describes how the price each successful bidder is awarded is determined. Typically, the design options can be distinguished by discriminatory and uniform pricing (Anatolitis and Welisch 2017). In discriminatory auctions, each bidder receives exactly the bid they submitted. In the case of a single-item auction, we refer to this concept as the “first price” rule, while in multi-item auctions, this concept is called “pay as bid” pricing. The alternative is uniform pricing, in which the bid for each awarded bidder is equal. In single-item auctions, the concept is referred to as the “second price” rule, as the awarded bidder, typically the bidder with the lowest bid price, receives the submitted price of the second lowest bidder. In multiitem auctions, the concept is referred to as “uniform pricing” or “pay as cleared”. Typically, the awarded bidders receive either the highest awarded or lowest non-awarded bid price. While in theory both pricing rules should result in the same outcome, uniform pricing provides incentives to bidders to submit their true costs. Pay-as-bid pricing can lead to higher bid prices, if bidders incorporate a higher mark up on their LCOE, thus harming support cost efficiency. Nevertheless, uniform pricing can lead to underbidding, as bidders might bid below their LCOE, hoping for higher bids to set a sufficient level of support. This underbidding might harm the realisation rates and thus the effectiveness. Pay-as-bid might be easier to comprehend, and can attract smaller bidders, supporting actor diversity. Floor price A floor price defines the minimum acceptable bid. Submitted bid prices under the floor price are not considered (Cassetta et al. 2017; Yalılı et al. 2020). As the floor price prohibits bidders from submitting a lower price as they might have intended, the floor price tends to harm support cost efficiency. Consequently, static efficiency is harmed as well, since in some cases, all of the bidders might submit the floor price, thus rendering the allocation mechanism ineffective, as the auction cannot guarantee that bidders with the lowest generation costs are actually awarded. On the other hand, the floor price prevents bidders from underbidding, thus reducing the likelihood of too aggressive bidding. Consequently, the awarded projects’ realisation probability increases, which positively affects the effectiveness. Actor diversity is increased by a floor price as well, as smaller bidders might have a higher chance of being awarded, if the floor price is not set too low.

Renewable energy auctions 

405

Ceiling price A ceiling price defines the maximum acceptable bid. Submitted bid prices above the ceiling price are not considered in the auction procedure (del Río 2017b). Implementing a ceiling price can ensure that submitted bids are not too high and thus ensure support cost efficiency. Nevertheless, if the ceiling price is set too low, some bidders might be excluded from the auction; thus the level of competition might decrease. This can lead to higher awarded prices, thus harming support cost efficiency. Similarly, bidders might orientate their bids towards the ceiling price; thus if set too high, the ceiling price might again harm support cost efficiency. Furthermore, a low ceiling price might exclude smaller actors (with comparably higher costs) from participating in the auction, thus harming actor diversity. Automatic volume adjustment Automatic volume adjustment, sometimes referred to as “endogenous rationing” (Hanke and Tiedemann 2020), aims at ensuring sufficient competition in an auction round. In general, if an auction is undersubscribed, the auctioneer reduces the original auctioned volume, to have sufficient competition in the auction (Anatolitis 2020; Ehrhart et al. 2020). The auctioneer has two design options: (1) bidders need to first submit an application for participating in the auction. If the auction is undersubscribed, the auctioneer reduces the original auctioned volume, so that the final auctioned volume is surpassed by the submitted bids’ capacity. Alternatively, (2) the auctioneer reduces the original auctioned volume after the actual auction procedure, i.e., fewer bidders are awarded than initially anticipated. The auctioneer typically declares the desired level of competition before the auction, e.g., that the awarded capacity needs to be surpassed by the submitted bids’ capacity by at least 75% (see Greek RE auctions (Anatolitis 2020)). As the automatic volume adjustment ensures competition in the auctions, which is usually a prerequisite for low bid prices, it is favourable for support cost efficiency. Otherwise, bidders would submit bid prices at the ceiling price if the auction were undersubscribed. Static efficiency might be harmed, as the higher allocation risk might lead to cost of capital and thus LCOE. Nevertheless, the automatic volume adjustment harms dynamic efficiency, as technologies with currently higher generation costs might never be awarded and, thus, future reductions in generation costs cannot be realised. Moreover, effectiveness is harmed, as the original awarded volumes are reduced, which might lead to a non-achievement of the RE targets. Actor diversity is harmed as well, as smaller actors (who tend to have higher LCOE and thus bid prices) have lower chances of being awarded, as the automatic volume adjustment ensures that some bidders are not awarded in the auction. Selection criteria The selection criteria define how bidders are awarded (del Río and Kiefer 2021). In price-only auctions, auctioneers award bidders with the lowest submitted bid price. In multi-criteria auctions, further criteria are taken into account when bidders are awarded (e.g., environmental impact, impact on industry and jobs). In price-only auctions, bidders tend to focus more on submitting low bid prices than in multicriteria auctions. Furthermore, if additional (non-price) criteria are included in the selection process, bidders with the lowest submitted bid prices (and typically lowest generation costs) might not necessarily be awarded. Consequently, multi-criteria auctions tend to harm support cost and static efficiency. Multi-criteria auctions might be more favourable for actor diversity, since smaller actors might not be able to compete with regard to their bid prices but might

406  Handbook on the economics of renewable energy

potentially be able to compete on the other criteria. On the other hand, additional criteria might increase the transaction costs for smaller players, thus making their participation less likely. Depending on whether the objectives are included as criteria, multi-criteria auctions can be favourable for dynamic efficiency, green growth and security of supply. Bonus/malus The auctioneer can introduce a bonus for certain types of projects/actors to increase their chances of being awarded or a malus to decrease the chances of certain projects/actors (IRENA 2019; Kreiss et al. 2021). Bonuses either decrease the prices in the award procedure, without altering the actual awarded price, or they are an actual top-up payment for certain bidders, so they can bid lower in the auctions. Maluses work the other way around: they either increase the prices of certain actors in the auction procedure or reduce the actual payments after the auction. Both bonuses and maluses harm both support cost and static efficiency, as it is not necessarily the projects with the lowest bid prices and lowest generation costs that are awarded. Nevertheless, they tend to increase actor diversity and can increase dynamic efficiency, if the supported technologies become cheaper in the future. Last awarded bid The auctioneer needs to decide how to deal with a bid surpassing the auctioned volume (in multi-item auctions). Four design options exist: (1) the bid is awarded entirely, thus surpassing the auctioned volume, (2) the bid is awarded up to the auctioned volume (for instance only part of the project’s capacity is awarded), (3) the bid is not awarded at all, thus the auctioned volume is not entirely awarded, and (4) the next highest bid is awarded, as long as it fits in the residual auctioned volume. Option 1 is favourable for effectiveness, as the auctioned volume is surpassed but, since the last project typically entails a comparably high bid price, this option might harm support cost efficiency. Option 2 is suitable for effectiveness, and to a certain extent for support cost efficiency (as the entire volume of the project with a high bid price is not supported), but it is more difficult to implement and entails higher administrative efforts. Option 3 is favourable for support cost efficiency, as the comparably higher bid price is not awarded, yet effectiveness might suffer, since not all the auctioned volume is awarded. Lastly, while option 4 might be good for effectiveness, as the auctioned volume is awarded as closely as possible, this option is the worst for support cost efficiency, because projects with relatively high bid prices (but a low bid volume) might be awarded. Support and contract design Remuneration scheme The remuneration scheme determines how the support to the bidders is paid out. Several design options exist: feed-in tariffs (FITs) consist of a guaranteed price generators receive for their generated electricity – independent of the electricity market price. In contrast, under FIPs, generators are required to sell their generated electricity on the market themselves. Subsequently, they receive a support payment on top of their electricity market revenues. Different types of feed-in premiums exist, each with a different market price exposure. A fixed FIP is usually set once and remains unchanged afterwards and, thus, is a constant payment on top of the electricity market revenues. Support payments under a sliding FIP (also referred to as contractsfor-difference) change over time, as they are typically calculated as the difference between

Renewable energy auctions 

407

the awarded strike price in the auctions and the electricity market revenues. For the electricity market revenues, a market value is usually calculated, which is the weighted average electricity market price over a certain period of time. If market values are above the strike price, generators can retain this additional revenue under a one-sided sliding FIP. Under a two-sided sliding FIP, generators need to transfer those to the government. We will refer to the two-sided sliding FIP as “Contract-for-Difference (CfD)”. In contrast to the aforementioned generationbased remuneration schemes, the investment grant is a capacity-based payment, typically paid out at the beginning of the project to cover (part of) the investment expenditures. The main difference between those remuneration schemes is the extent to which generators are exposed to market prices. As with all other risks, they tend to increase the cost of capital, thus leading to higher LCOE, which in turn tend to increase the bid prices (Đukan and Kitzing 2023). Under a FIT, bidders receive a constant payment for each generated unit of electricity without any market price exposure. Thus, a FIT entails no price risk. The one-sided sliding FIP can be riskier under certain circumstances than the CfD, especially if bidders submit bid prices under their generation costs, expecting additional electricity market revenues. The fixed FIP represents usually a low share of the electricity market revenues, which is why it exposes generators to quite a lot of market price risk. Investment grants constitute a capacity-based remuneration scheme, meaning generators are fully exposed to market price risks. Consequently, we can state that the LCOE under a FIT should be the lowest, while the LCOE under a one-sided sliding FIP are at least as high as under a CfD, followed by the LCOE under a fixed FIP. The LCOE under an investment grant should be the highest. Based on this relation, we can derive a first indication of the submitted bid prices. In a competitive auction, we can assume bidders submit their LCOE under a FIT scheme. In contrast, if average market values are expected to be higher than the LCOE, bidders under a one-sided sliding FIP have the incentive to bid below their LCOE, even down to zero (see, e.g., Kreiss et al. 2017a; Neuhoff et al. 2018). Under the same conditions, bidders under CfDs will submit a price somewhere between their LCOE and the expected market value. Thus, expecting sufficiently high market values, we expect to observe the lowest bids under a one-sided sliding FIP, then FITs, then CfDs (Neuhoff et al. 2018). In the case of low expected market values in the future, we presume bids to be lowest under a FIT, while CfDs and one-sided sliding FIP would lead to the same bid prices. Thus, it becomes clear that the exact relation between bid prices under these three schemes is highly dependent on expectations of future market values. A direct comparison of the FIT, the one-sided sliding FIP and the CfD with the fixed FIP and the investment grant is difficult. However, we can state that the support cost efficiency under a fixed FIP and the investment grant depends on the expected market values, as well: high expected market values lead to lower bids, as the market revenues might be sufficient to cover generation costs, and vice versa. The effectiveness is highest under a FIT, followed by a CfD, a one-sided sliding FIP, a fixed FIP and an investment grant, following the price risk for bidders and the likelihood of underbidding and thus the “winner’s curse”. In terms of actor diversity, we can expect the FIT to perform best, as the direct marketing of electricity, which is required by a one-sided sliding FIP and a CfD, poses an additional challenge for small actors. Then, the CfD will perform at least as well as the one-sided sliding FIP, as bidders have the incentive to bid close to their true costs. Under a fixed FIP and an investment grant, smaller actors might have difficulties quantifying their future revenues, in addition to having to bear a higher market price risk. Nevertheless, it should be noted that the short- and long-term market integration of RE projects can be an objective, as well. Thereby, the one-sided sliding FIP, the fixed FIP and

408  Handbook on the economics of renewable energy

the investment grant perform best compared to the other remuneration schemes, with the FIT providing very few to no incentives for market integration. Closely related to the remuneration scheme is the calculation of the reference market value, which is used to determine the actual support paid out to the awarded bidders under a onesided sliding FIP and a CfD (Anatolitis and Klobasa 2019). The reference market value serves as a proxy for the electricity market revenues of RE projects and is calculated as the average over a certain period of time. Typically, the reference market value entails a temporal focus, the reference period, e.g., hourly, daily, monthly or yearly, over which the market revenues are averaged. Moreover, there is a technological focus, as well, meaning whether the average is weighted with the electricity generation by a certain RE technology to account for seasonal changes and cannibalisation effects. In general, the longer the reference period and the more technologies are included in the reference market value calculation, the higher the risk exposure for RE projects. This increased risk can potentially lead to higher costs of capital, which then lead to a higher LCOE. This increase in the LCOE can then potentially lead to higher bid prices and thus an increased need for support, consequently reducing support cost efficiency. Support duration The support duration defines the time period an awarded bidder is entitled to the support payments, usually defined in years (Szabó et al. 2021). The support periods typically vary between 12 and 25 years. In some instances, the support duration is further limited by a total amount of generated electricity, e.g., as in the current Spanish auction scheme (del Río and Menzies 2021). In principle, the longer the support duration, the lower the price exposure for RE generators and, thus, the lower the bid prices. Nevertheless, a shorter support duration reduces the period during which support payments accrue. Furthermore, the effect of the support duration on bid prices depends on whether the remuneration scheme is a one-sided FIP or a two-sided CfD. Therefore, the effect of the length of support duration on support cost efficiency is rather ambiguous. However, a longer support duration provides more security and easier access to financing for smaller actors, thus increasing actor diversity. Flexibility Policymakers can provide flexibility to bidders with regard to their awarded projects’ volumes (Anatolitis et al. 2022; IRENA 2015). Awarded bidders have the possibility to change their awarded volume slightly, i.e., within certain boundaries. For instance, awarded bidders in Germany can reduce their project’s capacity by 5% (EEG 2021). This flexibility reduces the risk for bidders and thus the risk premiums, leading to lower bid prices and thus higher support cost efficiency. Nevertheless, the flexibility might harm effectiveness if a large number of bidders make use of the flexibility and reduce their projects’ volumes. Realisation period The realisation period defines a deadline by which the awarded projects need to be realised (del Río and Kiefer 2021). In some instances, further milestones until the end of the deadline are defined, such as a final investment decision or signature of the contracts with the equipment suppliers. Typically, if the realisation period is not met, penalties accrue and/or the support agreement is withdrawn. In general, a longer realisation period can decrease the risk of the associated penalties for bidders, thus increasing support cost efficiency. Longer realisation periods encourage bidders

Renewable energy auctions 

409

to speculate on decreasing technology costs and thus to submit low bid prices. Nevertheless, this entails the risk that projects are not built if the technology costs do not decrease as expected, thus harming effectiveness. A longer realisation period can also be favourable for actor diversity, as smaller players might need a longer time to realise their projects. Penalties Policymakers can introduce penalties to safeguard the realisation of awarded projects (Kreiss et al. 2017b). Penalties accrue if awarded projects are not realised by a certain deadline. The penalties can accrue gradually, if policymakers aim at penalising delays less severely. Penalties increase the risk for project developers and thus the risk premiums, thus leading to potentially higher bid prices, harming support cost efficiency. High penalties can even impede bidders from participating in the auction, thus decreasing the level of competition. In general, penalties have a positive impact on effectiveness as they ensure a high realisation rate of projects. Nevertheless, if they are set too low (or have not been implemented), they might encourage underbidding and thus endanger the realisation of projects. Furthermore, if the penalties are set too high, smaller actors might be excluded from participating in the auction, due to challenges with accessing financing, which might harm actor diversity.

ZERO-PRICE BIDS IN OFFSHORE WIND AUCTIONS Regarding support cost efficiency, the most notable outcome was the zero-price bids: in several offshore wind auctions, the winning bids amounted to 0 EUR/MWh. These zeroprice bids occurred first in the transitional German offshore wind auctions in 2017 and 2018. Nevertheless, in the regular German offshore wind auctions in 2021, and in the Danish Thor auction in 2021 (Jansen et al. 2022), they were observed as well. The Dutch offshore wind auctions from 2017 do not foresee any support payments at all, meaning zero-price bids are already implemented in the auction design. Thus, bidders participate in the auction to receive the right to realise an offshore wind farm on a certain site, competing based on a set of nonprice criteria, for instance, the environmental impact of the offshore wind farm or the level of innovative solutions (Netherlands Enterprise Agency 2022). In the past, as generation costs of offshore wind projects were typically higher than the electricity market prices, governmental support was needed to make these projects profitable. With decreasing LCOEs and increasing electricity market prices, offshore wind projects became profitable even without (direct) financial support from the government. Thus, auctions gradually changed from being an instrument to allocate support to a mechanism that awards the right to build an already profitable project, which is, at least for offshore wind, not possible outside the support schemes. Combined with high levels of competition in the auctions, bidders have the incentive to forfeit any right to governmental support by submitting zero-price bids in order to increase their chances of being awarded, realise their project and receive the electricity market revenue (Kreiss et al. 2017a). Germany uses a one-sided sliding FIP in its RE auction design meaning that RE generators can retain these additional revenues if electricity market prices are above the awarded strike price (Anatolitis and Klobasa 2019). Thus, project developers internalise these additional revenues when calculating their bid price. If the (expected) electricity market revenues are above the project’s LCOE, the project becomes profitable even without support from the government. Thus, the right to build the project became the contested “good” of the auction.

410  Handbook on the economics of renewable energy

Since the projects were already profitable, the bidders decreased their submitted bid prices to 0 EUR/MWh to increase their chances of being awarded. This was especially relevant in the transitional offshore wind auctions, as the project developers lost their rights to their already pre-developed projects, increasing the pressure to be awarded. However, even in the first regular offshore wind auctions in Germany in 2021, several bidders submitted zero-price bids, indicating that project developers perceive their projects as profitable without governmental support payments. Similarly, the Danish Thor auction in 2021 received several zero-price bids, although a (two-sided) CfD was in place. Under a CfD, RE generators need to transfer the electricity market revenues above the awarded strike price to the government, making zero-price bids unfeasible in principle. Nevertheless, in the Thor auction, the Danish government implemented a payback cap, limiting the RE generators’ payback to the government. Consequently, the payback makes the remuneration scheme more similar to a one-sided sliding FIP with a negative bid component and less similar to a classic CfD. Subsequently, zero-price bids can emerge if the following conditions are met: 1. 2. 3. 4.

No alternative to realise the project outside of the support scheme. Auction-based allocation mechanism with high competition. High expected electricity market prices (above the LCOE). Electricity market revenues (above the awarded bid/strike price) can be retained.

It should be noted that zero-price bids do not only occur in offshore wind auctions, but in auctions for other technologies, as well. For instance, in the Portuguese PV auctions, negative fixed FIPs are foreseen, which constitute a fixed payment per generated unit of electricity from the RE generators to the government, reducing the electricity market revenues of the generators (del Río et al. 2019b). However, the RE auctions in Portugal are similar to the offshore wind auctions in other countries. As grid access is limited in Portugal, project developers rather compete to obtain the right to build their already profitable projects. As zero-price bids emerged in the offshore wind sector and are much more common there, we will focus in the following on zero-price bids in offshore wind auctions. While zero-price bids reflect the increased maturity and competitiveness of the offshore wind sector, they pose several challenges to policymakers. There are concerns that the award in offshore wind auctions might be perceived as merely an option to realise the project (“option bidding”). Projects with zero-price bids are fully exposed to market price risks, so they might not be realised if the electricity market price development does not materialise as expected. Furthermore, receiving several zero-price bids in the auction signals that developers might have a higher “willingness to pay” for the right to realise the project, which means that additional profits for the government/society are not realised. And lastly, in case of multiple zero-price bids in an auction, the auctioneer cannot differentiate between the different bids on price alone and thus often resorts to drawing lots, which is a rather ineffective and inefficient way of determining the winner. One question that emerges is how policymakers can deal with these zero-price bids in auctions. In the following, we present several potential approaches. While the focus is on the award procedure, we also discuss materialising the developers’ willingness to pay. Given this context, we distinguish between two types of approaches. The first type focuses on decreasing the likelihood of receiving zero-price bids in the auctions and includes:

Renewable energy auctions  ● ● ●

411

Deep grid connection costs Seabed lease contracts/auctions Contracts-for-differences

The second type focuses on the mechanism of selecting a winner if multiple zero-price bids have been submitted and includes: ● ● ●

Lottery Negative bid component Beauty contest/multi-criteria selection process

Approaches for Reducing the Likelihood of Zero-Price Bids Deep grid connection costs In contrast to the “shallow grid connection costs”, where costs for the grid connection are borne by the government/consumers, project developers need to bear the full costs under “deep grid connection costs”. Internalising these costs can make the project unprofitable without governmental support. Thus, incorporating these higher costs into the bid calculation can reduce the likelihood of zero-price bids. Denmark (since the Thor auction) and the UK have implemented deep grid connection costs in their offshore wind auctions (Jansen et al. 2022). This approach entails several advantages: if bidders internalise the grid connection cost and zero-price bids do not occur, the auctioneer can differentiate between the bid prices and award the bidder with the lowest bid price and thus most likely with the lowest generation costs. Therefore, this approach increases the likelihood of a support cost and static efficient outcome. In addition, this approach can be implemented relatively easily. And, lastly, the government can materialise the willingness to pay off project developers and thus reduce the windfall profits of project developers, even in the case of zero-price bids. Nevertheless, zero-price bids might still occur under deep grid connection costs, if the expected electricity market revenues are higher than the combination of LCOE and grid connection costs. Furthermore, potential challenges with regard to (efficient) grid planning might arise and cost allocation might arise. Seabed lease contracts/auctions Some countries, such as the UK and the US, have introduced auctions for allocating seabed leases (Jansen et al. 2022). Project developers with the highest willingness to pay for the seabed lease receive the right to develop an offshore wind project on a certain area on the seabed. Similar to the deep grid connection costs, auctioning the seabed leases increases bidders’ costs and thus reduces the likelihood of zero-price bids. If the land lease payment is not administratively set, the auction might lead to a support cost and static efficient outcome. Again, the government can materialise the willingness to pay off project developers for the seabed leases, which would otherwise be granted for free or for an administratively set fee. Thus, potential windfall profits of project developers can be reduced, even in cases of zero-price bids. Nevertheless, auctioning the seabed lease contracts can induce high transaction costs, as typically, an additional auction needs to be conducted. Zero-price bids might still occur, if the expected electricity market revenues are higher than the combined LCOE and seabed lease

412  Handbook on the economics of renewable energy

costs. Furthermore, the likelihood of the winner’s curse increases, as the bidder with the highest expectations of future electricity market prices will most likely bid the highest price and thus win the seabed lease auctions. If the electricity market prices remain below the expectations, the project might not be realised, which harms the scheme’s effectiveness. Contracts-for-difference Zero-price bids first emerged in Germany, where a one-sided sliding FIP has been in place. Under this remuneration scheme, plant operators receive support payments as long as the electricity market prices/market value of offshore wind is lower than the bid/strike price. If the electricity market prices are above the strike price, the operators do not receive any support payments but can keep this additional revenue. In contrast, in a (two-sided) CfD, plant operators have to pay back revenues above the strike price to the supporting body. Thus, zero-price bids are not feasible, since all the revenues from the electricity market would need to be paid back to the supporting body. A notable exception is the Thor auction in Denmark in 2021, in which several zero-price bids were submitted, although a CfD was auctioned. Zero-price bids were still feasible, as the overall payback to the government was capped at DKK 2.8 billion (roughly 375 million EUR), meaning revenues from the electricity market beyond this cap can be kept by the operator. Besides Denmark, France and the UK have CfDs in place (Jansen et al. 2022). Introducing CfDs most likely helps the auctioneer to select the winner effectively, since bidders can be distinguished based on their submitted bid prices (that are most likely not zeroprice bids). Furthermore, the auctioneer can award the bidder with the lowest bid price and thus potentially lowest generation costs, thus ensuring static efficiency. Whether the outcome is support cost efficient, as well, depends on the development of the electricity market prices. If they are above the strike price, the government will receive money from the awarded bidder, and thus have a better outcome compared to a zero-price bid under a one-sided sliding FIP. In contrast, if the electricity market prices remain below the awarded CfD strike price, the government needs to provide support to the awarded bidder, thus harming support cost efficiency. Introducing CfDs in an existing support scheme can be fairly simple and, most likely, a transparent selection based on bid prices is ensured. Again, the government can materialise the willingness to pay of project developers. Nevertheless, if the overall payback from project developers is capped, zero-price bids can still occur (as in the Thor auction in Denmark). Moreover, CfDs tend to harm the market integration of RE, compared to one-sided sliding FIP, since zero-price bids under a FIP are usually combined with long-term power purchase agreements (PPAs). Approaches for Selecting a Winner among Zero-Price Bids Lottery In the case of several zero-price bids, the auctioneer can select the winner randomly by conducting a lottery. This approach is followed, for instance, in Denmark and Germany. The biggest advantage of a lottery is that it is easily implemented and conducted. In general, it entails a low administrative effort, as well as low transaction costs. However, an efficient auction outcome is not ensured, as the project with the lowest LCOE might not be awarded. Furthermore, windfall profits for RE developers occur, as they might have a higher willingness to pay for the right to realise the project (beyond the zero-price bid).

Renewable energy auctions 

413

Lastly, a lottery might be regarded as a non-transparent selection procedure by bidders, which in turn increases the risk premiums. Negative bid component Another approach is asking bidders to accompany a zero-price bid with a negative bid component (this can be interpreted as a bid reflecting the “willingness to pay”), which is typically expressed as a lump sum. This negative bid component can be communicated as a contribution to the (shallow) grid connection or as the seabed lease costs. The auctioneer can select the winner based on the submitted negative bid component. The Netherlands has introduced such a negative bid component in their 2022 offshore wind auction, albeit it has capped the sum to 50 million EUR (Netherlands Enterprise Agency 2022). If the negative bid component is not capped, this approach allows an effective selection of bidders based on their submitted price. Assuming the bidder with the lowest generation cost can submit the highest negative bid component, the auction can lead to a static efficient outcome. Moreover, a support cost efficient outcome can be reached. By asking bidders to submit negative bid components, the government can materialise the willingness to pay off project developers and thus decrease their windfall projects. Nevertheless, since bidders with the most optimistic expectations of future electricity market prices will most likely bid the highest negative bid component, the likelihood that the winner’s curse will occur rises. Thus, in case of underbidding, effectiveness might suffer. Beauty contest/multi-criteria selection process In a multi-criteria selection process, bids are evaluated not only based on their price but also on other criteria. For instance, France uses this approach in its offshore wind auctions. These auctions can include, but are not restricted to, innovation, local participation or employment, environmental issues, energy delivered and others. In contrast, in a beauty contest, the selection procedure does not contain any price criterion, such as in the Dutch offshore wind auctions between 2017 and 2020 (Jansen et al. 2022). In most cases, both multi-criteria selection processes and beauty contests allow for an effective selection of winners. Furthermore, additional policy objectives can be included in the selection process (e.g., industrial development). Nevertheless, a static efficient auction outcome is not ensured, as the project with the lowest generation costs might not be awarded. However, this might not be the most important policy objective if this approach is followed. Potential windfall profits for RE developers might occur, as they might have a higher willingness to pay for the right to realise the project. Moreover, this approach entails a higher degree of complexity, which can increase the transaction costs and potentially decrease competition. Designing and weighting the selection criteria might be challenging. And, lastly, multi-criteria auctions/beauty contests might be regarded as non-transparent by bidders. Combination of multiple approaches Given the various advantages and drawbacks of each approach, countries usually employ a combination of the different approaches. Table 17.2 provides an overview of the different approaches/combinations implemented in Europe.

414  Handbook on the economics of renewable energy

Table 17.2  Overview of different approaches for dealing with zero-price bids in Europe Deep grid connection

Seabed lease contracts/auctions

Contracts-fordifference

DE DK

Negative bid component

Beauty contest/ multi-criteria selection process

X X

X

FR

X

X

NL UK

Lottery

X X

X

X

X

X

So far, only Germany and Denmark have had to resort to drawing lots, which is regarded as a rather ineffective and inefficient way to determine the auction winner. In other countries, zero-price bids either were avoided (as in the case of France and the UK) or had already been reflected in the auction design (as in the Netherlands). It should be noted that even if approaches that reduce the likelihood of zero-price bids are chosen, such as in the case of the Thor auction in Denmark, the details of the implementation matter and should thus be taken into account by policymakers.

THE FUTURE OF RE AUCTIONS Renewable energy auctions are expected to evolve in the near future. As already indicated, in the offshore wind sector, auctions are transforming from an instrument to allocate support to unprofitable projects to an instrument which allocates the right to build already profitable projects (Jansen et al. 2022). This trend can already be observed, as governments are gradually turning to seabed lease auctions, which aim at allocating competitively the site on which the offshore wind project can be realised, for instance in the UK or the US. Another example is the increased use of multi-criteria selection processes, often without any support payments, such as in the case of the Netherlands (Jansen et al. 2022). Moreover, RE technologies, especially onshore wind and PV, are becoming more mature and, therefore, are more comparable regarding their costs. Thus, we expect to see more multi-technology RE auctions in which these two technologies compete against each other in the near future. Furthermore, to ensure security of supply, RE auctions might start focusing more on technology combinations, such as PV or onshore wind projects with storage. These solutions have already been auctioned in several RE auctions, such as in the German innovation auctions (Sach et al. 2019) and in the Portuguese auctions (del Río et al. 2019b). While auctions have so far been mostly used to allocate governmental financial support to RE projects, we observe increased use by other actors in the RE sector. For instance, private actors, such as Google, have already started using auctions for procuring electricity from RE in the form of corporate PPAs (Google 2019). One further application for auctions that is currently being discussed is the support for (green) hydrogen (e.g., Kerres et al. 2022; Zheng et al. 2022). In contrast to RE auctions, it is still unclear whether the supply or the demand side should be supported. Moreover, the

Renewable energy auctions 

415

missing transport infrastructure for hydrogen makes the auction design much more complex compared to RE auctions. Nevertheless, the experience from RE auctions can be used, e.g., for designing the remuneration scheme or setting the penalties. Furthermore, auctions might be used to allocate support for the decarbonisation of the industry, e.g., through the allocation of carbon contracts-for-differences (Richstein and Neuhoff 2022; Rilling et al. 2022) or through competitively selecting projects in the EU’s Innovation Fund (Winkler et al. 2022). Additionally, auctions have already been used for allocating support for energy efficiency measures (Anatolitis and Schlomann 2022) and could play a role in the heating and cooling sector (Winkler et al. 2022). In the transport sector, auctions have already started to be used to support the implementation of charging infrastructure for electric vehicles (Winkler et al. 2022). Consequently, the experience and insights from RE auctions will be valuable for the design of future applications of auctions.

NOTES 1.

There is no clear distinction between “auctions” and “tenders” in the literature. In some instances, the literature refers to auctions in case the competitive bidding process relies only on the submitted prices, while in tenders, further criteria besides the price are taken into account. Nevertheless, in this chapter, we will use the term auctions for all competitive bidding processes. 2. One could argue that awarded prices do not necessarily reflect the actual support expenditures, especially in cases of sliding feed-in premiums. In this remuneration scheme, the actual support payments are calculated as the difference between the strike price, i.e. the awarded bid price, and a certain proxy for the electricity market revenues. Nevertheless, taking into account future market values/electricity market revenues is quite a difficult challenge. In addition, it is possible to assess the support cost efficiency as the awarded price as ceteris paribus, as the same RE project could have submitted a higher/lower price and still have the same market value.

REFERENCES Al-Sarihi, A., and N. Mansouri. (2022). ‘Renewable energy development in the Gulf Cooperation Council countries: Status, barriers, and policy options’. Energies, 15(5). https://www​.mdpi​.com​/1996​1073​/15​/5​/1923. Álvarez, F., and P. del Río. (2022). ‘Is small always beautiful? Analyzing the efficiency effects of size heterogeneity in renewable electricity auctions’. Energy Economics, 106, 105698. https://doi​.org​/10​ .1016​/j​.eneco​.2021​.105698. Anatolitis, V. (2020). ‘Auctions for the support of renewable energy in Greece’. AURES II Report D2.1-EL. http://aures2project​.eu​/wp​-content​/uploads​/2020​/03​/AURES​_II​_case​_study​_Greece​.pdf. Anatolitis, V., and M. Klobasa. (2019). ‘Impact of a yearly reference period on the sliding feed-in premium for onshore wind in Germany’. In: 16th International Conference on the European Energy Market (EEM), 1–7. https://doi.org/10.1109/EEM.2019.8916394. Anatolitis, V., and B. Schlomann. (2022). ‘Auctions for energy efficiency and the experience of renewable energy’. Odyssee-Mure Policy Brief. https://www​.odyssee​-mure​.eu​/publ​​icati​​ons​/p​​olicy​​ -brie​​f​/auc​​tions​​-ener​​gy​-ef​​ficie​​ncy​-r​​es​.pd​​f. Anatolitis, V., and M. Welisch. (2017). ‘Putting renewable energy auctions into action – An agent-based model of onshore wind power auctions in Germany’. Energy Policy, 110, 394–402. https://doi​.org​/10​ .1016​/j​.enpol​.2017​.08​.024. Anatolitis, V., A. Azanbayev, and A.-K. Fleck. (2022). ‘How to design efficient renewable energy auctions? Empirical insights from Europe’. Energy Policy, 166, 112982. https://www​.sciencedirect​ .com​/science​/article​/pii​/S0301421522002075.

416  Handbook on the economics of renewable energy

AURES II. (2022). ‘Auction database’. https://doi​.org​/10​.17632​/2h9jpzy3wx​.1. Batz Liñeiro, T., and F. Müsgens. (2023). ‘Evaluating the German onshore wind auction programme: An analysis based on individual bids’. Energy Policy, 172, 113317. https://doi​.org​/10​.1016​/j​.enpol​.2022​ .113317. Bayer, B. (2018). ‘Experience with auctions for wind power in Brazil’. Renewable and Sustainable Energy Reviews, 81, 2644–2658. https://doi​.org​/10​.1016​/j​.rser​.2017​.06​.070. Cassetta, E., U. Monarca, C. R. Nava, and L. Meleo. (2017). ‘Is the answer blowin’ in the wind (auctions)? An assessment of the Italian support scheme’. Energy Policy, 110, 662–674. https://doi​.org​/10​.1016​/j​ .enpol​.2017​.08​.055. Côté, E., M. Đukan, C. Pons-Seres de Brauwer, and R. Wüstenhagen. (2022). ‘The price of actor diversity: Measuring project developers’ willingness to accept risks in renewable energy auctions’. Energy Policy, 163, 112835. https://doi​.org​/10​.1016​/j​.enpol​.2022​.112835. del Río, P. (2016). ‘Auctions for renewable energy support in Portugal: Instruments and lessons learnt’. AURES Report D4.1-PT. http://aures2project​.eu​/2021​/07​/06​/auctions​-for​-renewable​-energy​-support​in​-portugal​-instruments​-and​-lessons​-learnt/. del Río, P. (2017a). ‘Auctions for renewable support in Mexico: Instruments and lessons learnt’. AURES Report D4.4-MX. http://aures2project​.eu​/2021​/07​/06​/auctions​-for​-renewable​-support​-in​-mexico​instruments​-and​-lessons​-learnt/. del Río, P. (2017b). ‘Designing auctions for renewable electricity support. Best practices from around the world’. Energy for Sustainable Development, 41, 1–13. https://doi​.org​/10​.1016​/j​.esd​.2017​.05​.006. del Río, P. (2017c). ‘Designing auctions for renewable electricity support: the case of Spain’. Renewable Energy Law and Policy Review, 8(2), 23–37. https://www​.elgaronline​.com​/view​​/jour​​nals/​​relp/​​8​-2​/r​​ elp​.2​​017​.0​​2​.02.​​xml. del Río, P. (2018). ‘An analysis of the design elements of the third renewable energy auction in Spain’. Renewable Energy Law and Policy Review, 8(3), 17–30. https://doi​.org​/10​.4337​/relp​.2018​.03​.03. del Río, P. (2019). ‘Auctions for the support of renewable energy in Mexico’. AURES II Report D2.1-MX. http://aures2project​.eu​/wp​-content​/uploads​/2019​/12​/AURES​_II​_case​_study​_ Mexico​.pdf. del Río, P., and E. Cerdá. (2014). ‘The policy implications of the different interpretations of the costeffectiveness of renewable electricity support’. Energy Policy, 64, 364–372. https://doi​.org​/10​.1016​/j​ .enpol​.2013​.08​.096. del Río, P., and C. Menzies. (2021). ‘Auctions for the support of renewable energy in Spain’. AURES II Report D2.1-ES. http://aures2project​.eu​/wp​-content​/uploads​/2021​/10​/AURES​_II​_case​_study​_ Spain​ .pdf. del Río, P., and C. Kiefer. (2021). ‘Analysing patterns and trends in auctions for renewable electricity’. Energy for Sustainable Development, 62, 195–213. https://doi​.org​/10​.1016​/j​.esd​.2021​.03​.002. del Río, P., and C. Kiefer. (2023). ‘Academic research on renewable electricity auctions: Taking stock and looking forward’. Energy Policy, 173, 113305. https://doi​.org​/10​.1016​/j​.enpol​.2022​.113305. del Río, P., and P. Linares. (2014). ‘Back to the future? Rethinking auctions for renewable electricity support’. Renewable and Sustainable Energy Reviews, 35, 42–56. https://doi​.org​/10​.1016​/j​.rser​.2014​ .03​.039. del Río, P., S. Steinhilber, and F. Wigand. (2015). ‘Assessment criteria for RES-E auctions’. http:// aures2project ​ .eu ​ / 2021​ / 07​/ 06 ​ /assessment​ - criteria ​ -for​- res ​ - e ​ - auctions/ ​ ? utm ​ _ source ​ = rss ​ & utm ​ _ medium​=rss​&utm ​_c​​ampai​​gn​=as​​sessm​​ent​-c​​riter​​ia​-fo​​r​-res​​-e​-au​​ction​​s. del Río, P., C. Kiefer, J. Winkler, and V. Anatolitis. (2019a). ‘Auctions for the support of renewable energy from Concentrated Solar Power (CSP)’. AURES II Report D2.1-CSP. http://aures2project​.eu​/ 2020​/02​/05​/auctions​-for​-the​-support​-of​-renewable​-energy​-from​-concentrated​-solar​-power​-csp/. del Río, P., H. Lucas, B. Dézsi, and A. Diallo. (2019b). ‘Renewable electricity auctions in Portugal’. AURES II Report D2.1-PT. Diniz, B. A., A. Szklo, M. T. Tolmasquim, and R. Schaeffer. (2023). ‘Delays in the construction of power plants from electricity auctions in Brazil’. Energy Policy, 175, 113467. https://doi​.org​/10​.1016​ /j​.enpol​.2023​.113467. Đukan, M., and L. Kitzing. (2021). ‘The impact of auctions on financing conditions and cost of capital for wind energy projects’. Energy Policy, 152, 112197. https://doi​.org​/10​.1016​/j​.enpol​.2021​.112197. Đukan, M., and L. Kitzing. (2023). ‘A bigger bang for the buck: The impact of risk reduction on renewable energy support payments in Europe’. Energy Policy, 173, 113395. https://doi​.org​/10​.1016​/j​ .enpol​.2022​.113395.

Renewable energy auctions 

417

EEG. (2021). ‘Erneuerbaren-Energien-Gesetz 2021’. https://www​.gesetze​-im​-internet​.de​/eeg​_2014​/ BJNR106610014​.html. Ehrhart, K.-M., A.-K. Hanke, V. Anatolitis, and J. Winkler. (2019). ‘Auction-theoretic aspects of crossborder auctions’. AURES II Report D6.2. http://aures2project​.eu​/wp​-content​/uploads​/2020​/02​/ MultiAuctions​_final​_anv​.pdf. Ehrhart, K.-M., A.-K. Hanke, and M. Ott. (2020). ‘A small volume reduction that melts down the market: Auctions with endogenous rationing’. ZEW Discussion Paper, 20(14). https://ftp​.zew​.de​/pub​ /zew​-docs​/dp​/dp20014​.pdf. Eicke, L., and S. Weko. (2022). ‘Does green growth foster green policies? Value chain upgrading and feedback mechanisms on renewable energy policies’. Energy Policy, 165, 112948. https://doi​.org​/10​ .1016​/j​.enpol​.2022​.112948. European Commission. (2014). ‘Communication from the commission – Guidelines on state aid for environmental protection and energy 2014–2020’. Official Journal of the European Union (2014/C 200/01). Fitch-Roy, O. (2015). ‘Auctions for renewable support in California: Instrument and lessons learnt’. AURES Report D4.1-CAL. http://aures2project​.eu​/2021​/07​/06​/auctions​-for​-renewable​-energy​ -support​-in​-california​-instruments​-and​-lessons​-learnt/. Fitch-Roy, O. W., D. Benson, and B. Woodman. (2019). ‘Policy instrument supply and demand: How the renewable electricity auction took over the world’. Politics and Governance, 7(1), 81–91. https://doi​ .org​/10​.17645​/pag​.v7i1​.1581. Fleck, A.-K., and V. Anatolitis. (2023). ‘Achieving the objectives of renewable energy policy – Insights from renewable energy auction design in Europe’. Energy Policy, 173, 113357. https://doi​.org​/10​.1016​ /j​.enpol​.2022​.113357. Förster, S. (2016). ‘Small-scale PV auctions in France: Instruments and lessons learnt’. AURES Report D4.1-FR.  http://aures2project​.eu ​/2021​/07​/06​/auctions​-for​-renewable ​- energy​-support​-in​-france ​instruments​-and​-lessons​-learnt/. Förster, S., and A. Amazo. (2016). ‘Auctions for renewable energy support in Brazil: Instruments and lessons learnt’. AURES Report D4.1-BRA. http://aures2project​.eu​/2021​/07​/06​/auctions​-for​-renewable​energy​-support​-in​-brazil​-instruments​-and​-lessons​-learnt/. Google. (2019). ‘Case study: Accelerating renewable energy purchasing through auctions’. https:// services​.google​.com ​/fh ​/files​/misc​/case​-study​-renewable​-energy​-auctions​.pdf. Grashof, K. (2019). ‘Are auctions likely to deter community wind projects? And would this be problematic?’ Energy Policy, 125, 20–32. https://doi​.org​/10​.1016​/j​.enpol​.2018​.10​.010. Grashof, K. (2021). ‘Who put the hammer in the toolbox? Explaining the emergence of renewable energy auctions as a globally dominant policy instrument’. Energy Research & Social Science, 73, 101917. https://doi​.org​/10​.1016​/j​.erss​.2021​.101917. Greg, Buckman, Jon Sibley, and Megan Ward. (2019). ‘The large-scale feed-in tariff reverse auction scheme in the Australian Capital Territory 2012, to 2016’. Renewable Energy, 132, 176–85. https://​ www​.sciencedirect​.com​/science​/article​/pii​/S0960148118309637. Haas, R., G. Resch, C. Panzer, S. Busch, M. Ragwitz, and A. Held. (2011). ‘Efficiency and effectiveness of promotion systems for electricity generation from renewable energy sources – Lessons from EU countries’. Energy, 36(4), 2186–2193. Haelg, L. (2020). ‘Promoting technological diversity: How renewable energy auction designs influence policy outcomes’. Energy Research & Social Science, 69, 101636. https://doi​.org​/10​.1016​/j​.erss​.2020​ .101636. Hanke, A.-K., and S. Tiedemann. (2020). ‘How (not) to respond to low competition in renewable energy auctions’. AURES II Policy Brief. http://aures2project​.eu​/wp​-content​/uploads​/2020​/06​/AURES​_II​ _Policy​_Brief​_ End​_Rationing​.pdf. Hansen, U. E., I. Nygaard, M. Morris, and G. Robbins. (2020). ‘The effects of local content requirements in auction schemes for renewable energy in developing countries: A literature review’. Renewable and Sustainable Energy Reviews, 127, 109843. Hastings-Simon, S., A. Leach, B. Shaffer, and T. Weis. (2022). ‘Alberta’s renewable electricity program: Design, results, and lessons learned’. Energy Policy, 171, 113266. https://doi​.org​/10​.1016​/j​.enpol​.2022​ .113266. Haufe, M.-C., and K.-M. Ehrhart. (2018). ‘Auctions for renewable energy support – Suitability, design, and first lessons learned’. Energy Policy, 121, 217–224. https://doi​.org​/10​.1016​/j​.enpol​.2018​.06​.027.

418  Handbook on the economics of renewable energy

Held, A., M. Ragwitz, M. Gephart, E. de Visser, and C. Kleßmann. (2014). ‘Design features of support schemes for renewable electricity’. Energy Regulators Regional Association. Howlett, M. (2009). ‘Governance modes, policy regimes and operational plans: A multi-level nested model of policy instrument choice and policy design’. Policy Sciences, 42(1), 73–89. https://doi​.org​/ 10​.1007​/s11077​- 009​-9079​-1. Huberty, M., H. Gao, J. Mandell, and J. Zysman. (2011). ‘Shaping the green growth economy: A review of the public debate and the prospects for green growth’. Green Growth Leaders. IRENA. (2015). ‘Renewable energy auctions: A guide to design’. IRENA. (2019). ‘Renewable energy auctions: Status and trends beyond price’. Jakob, M., P. Noothout, F. von Bluecher, and C. Klessmann. (2019). ‘Auctions for the support of renewable energy in the Netherlands: Results and lessons learnt’. AURES II Report D2.1-NL. http:// aures2project​.eu​/wp​-content ​/uploads​/2019​/12​/AURES​_II​_case​_study​_ Netherlands​.pdf. Jansen, M., P. Beiter, I. Riepin, F. Müsgens, V. J. Guajardo-Fajardo, I. Staffell, B. Bulder, and L. Kitzing. (2022). ‘Policy choices and outcomes for offshore wind auctions globally’. Energy Policy, 167, 113000. Kerres, P., F. Wigand, F. von Blücher, C. Klessmann, V. Anatolitis, L. Zheng, and J. Winkler. (2022). ‘Auctions for the support of green hydrogen’. AURES II Policy Brief. http://aures2project​.eu​/wp​content​/uploads​/2022​/04​/AURES​_II​_ Policy​_Brief​_hydrogen​_auctions​.pdf. Kitzing, L., C. Mitchell, and P. E. Morthorst. (2012). ‘Renewable energy policies in Europe: Converging or diverging?’. Energy Policy, 51, 192–201. https://doi​.org​/10​.1016​/j​.enpol​.2012​.08​.064. Kitzing, L., V. Anatolitis, O. Fitch-Roy, C. Klessmann, J. Kreiß, P. del Río, F. Wigand, and B. Woodman (2019). ‘Auctions for Renewable Energy Support: Lessons Learned in the AURES Project’. IAEE Energy Forum (Third Quarter 2019), 11–14. https://​www​.iaee​.org​/en​/p​​ublic​​ation​​s​/new​​slett​​erdl.​​aspx?​​ id​=80​​9. Kitzing, L., O. Fitch-Roy, M. Islam, and C. Mitchell. (2020). ‘An evolving risk perspective for policy instrument choice in sustainability transitions’. Environmental Innovation and Societal Transitions, 35, 369–382. https://doi​.org​/10​.1016​/j​.eist​.2018​.12​.002. Kreiss, J., K.-M. Ehrhart, and A.-K. Hanke. (2017a). ‘Auction-theoretic analyses of the first offshore wind energy auction in Germany’. Journal of Physics: Conference Series, 926, 12015. https://doi​.org​ /10​.1088​/1742​- 6596​/926​/1​/012015. Kreiss, J., K.-M. Ehrhart, and M.-C. Haufe. (2017b). ‘Appropriate design of auctions for renewable energy support‐Prequalifications and penalties’. Energy Policy, 101, 512–520. https://doi​.org​/10​.1016​ /j​.enpol​.2016​.11​.007. Kreiss, J., K.-M. Ehrhart, M.-C. Haufe, and E. R. Soysal. (2021). ‘Different cost perspectives for renewable energy support: Assessment of technology-neutral and discriminatory auctions’. Economics of Energy & Environmental Policy, 10(1). https://doi​.org​/10​.5547​/2160​-5890​.10​.1​.jkre. Kruger, W., and A. Eberhard. (2018). ‘Renewable energy auctions in sub-Saharan Africa: Comparing the South African, Ugandan, and Zambian Programs’. Wiley Interdisciplinary Reviews: Energy and Environment, 7(4), e295. https://doi​.org​/10​.1002​/wene​.295. Melliger, M. (2022). ‘Quantifying technology skewness in European multi-technology auctions and the effect of design elements and other driving factors’. Energy Policy, 175, 113504. https://doi​.org​/10​ .1016​/j​.enpol​.2023​.113504. Menzies, C., and M. Marquardt. (2019). ‘Auctions for the support of renewable energy in Alberta, Canada: Main results and lessons learnt’. AURES II Report D2.1-CA. Mitchell, C. (1995). ‘The renewables NFFO’. Energy Policy, 23(12), 1077–1091. https://doi​.org​/10​.1016​ /0301​- 4215(95)00123-9. Mitchell, C. (2000). ‘The Enland and Wales non-fossil fuel obligation: History and lessons’. Annual Review of Energy and the Environment, 25. https://doi​.org​/10​.1146​/annurev​.energy​.25​.1​.285. Mora, D., L. Kitzing, E. R. Soysal, S. Steinhilber, P. del Río, F. Wigand, C. Klessmann, S. Tiedemann, A. L. A. Blanco, and M. Welisch. (2017a). ‘Auctions for renewable energy support-Taming the beast of competitive bidding’. AURES Report D9.2. http://aures2project​.eu​/2021​/07​/06​/auctions​-for​ -renewable​-energy​-support​-taming​-the​-beast​-of​-competitive​-bidding/. Mora, D., M. Islam, E. R. Soysal, L. Kitzing, A. L. A. Blanco, S. Förster, S. Tiedemann, and F. Wigand (2017b). ‘Experiences with auctions for renewable energy support’. In: 2017 14th International Conference on the European Energy Market (EEM), pp. 1–6.

Renewable energy auctions 

419

Netherlands Enterprise Agency. (2022). ‘Hollandse Kust (west) wind farm zone: Appendix A: Applicable law part of project and site description’. https://offshorewind​.rvo​.nl​/file​/download​/c85ef7b0​- 6229​4055​-964b​-9667164e4624​/ hkw​_20220413​_psd​_appendix​-a​.pdf. Neuhoff, K., N. May, and J. C. Richstein. (2018). ‘Renewable energy policy in the age of falling technology costs’. DIW Berlin Discussion Paper (1746). https://​www​.diw​.de​/de​/d​​iw​_01​​.c​.59​​4403.​​de​/ pu​​blika​​tione​​n ​/dis​​kussi​​onspa​​piere​​/2018​​_1746​​/rene​​wable​​_ener​​gy​_po​​licy_​​in​_th​​e​_ age​​_of​_f​​allin​​g ​_tec​​ hnolo​​gy​_co​​sts​.h​​tml. Noothout, P., and T. Winkel. (2016). ‘Auctions for renewable energy support in the Netherlands: Instruments and lessons learnt’. AURES Report D4.1-NL. Paravantis, J. A., and N. Kontoulis. (2020). ‘Energy security and renewable energy: A geopolitical perspective’. In: M. Al Qubeissi (Ed.), Renewable Energy: Resources, Challenges and Applications. IntechOpen. Probst, B., V. Anatolitis, A. Kontoleon, and L. D. Anadón. (2020). ‘The short-term costs of local content requirements in the Indian solar auctions’. Nature Energy, 5, 842–850. https://doi.org/10.1038/s41560020-0677-7. r2b energy consulting GmbH. (2019). ‘Definition and monitoring of security of supply on the European electricity markets’. Federal Ministry of Economics and Energy. Ragwitz, M., and S. Steinhilber. (2014). ‘Effectiveness and efficiency of support schemes for electricity from renewable energy sources’. Wiley Interdisciplinary Reviews: Energy and Environment, 3(2), 213–229. https://doi​.org​/10​.1002​/wene​.85. REN21. (2022). ‘Renewables 2022 global status report’. https://​www​.ren21​.net​/wp​-c​​onten​​t ​/upl​​oads/​​ 2019/​​05​/GS​​R2022​​_ Full​​_Repo​​r t​.pd​​f. Resch, G., J. Geipel, and F. Schöniger. (2022). ‘Case cooperation with Austria: Briefs on Austrian case cooperation’. AURES II Report D2.4-AT. http://aures2project​.eu​/2022​/05​/02​/case​-cooperation​-with​ -austria/​?utm ​_source​=rss​&utm ​_medium​=rss​&utm ​_c​​ampai​​gn​=ca​​se​-co​​opera​​tion-​​with-​​austr​​ia. Richstein, J. C., and K. Neuhoff. (2022). ‘Carbon contracts-for-difference: How to de-risk innovative investments for a low-carbon industry?’ iScience, 25(8), 104700. Rilling, A., V. Anatolitis, and L. Zheng. (2022). ‘How to design carbon contracts for difference – A systematic literature review and evaluation of design proposals’. In: 18th International Conference on the European Energy Market (EEM). Sach, T., B. Lotz, and F. von Bluecher. (2019). ‘Auctions for the support of renewable energy in Germany: Main results and lessons learnt’. AURES II Report D2.1-DE. http://aures2project​.eu​/wp​-content​/ uploads​/2020​/04​/AURES​_II​_case​_study​_Germany​_v3​.pdf. Steinhilber, S. (2016a). ‘Auctions for renewable energy support in Ireland: Instruments and lessons learnt’. AURES Report D4.1-IE. http://aures2project​.eu​/wp​-content​/uploads​/2021​/07​/pdf3​_ireland​.pdf. Steinhilber, S. (2016b). ‘Onshore wind concessions in China: Instruments and lessons learnt’. AURES Report D4.1-CN. http://aures2project​.eu​/2021​/07​/06​/auctions​-for​-renewable​-energy​-support​-in​ -china​-instruments​-and​-lessons​-learnt/. Szabó, L., M. Bartek-Lesi, A. Diallo, B. Dézsi, V. Anatolitis, and P. del Río. (2021). ‘Design and results of recent renewable energy auctions in Europe’. Papeles de Energía, 13. https://​www​.funcas​.es​/ revista ​/papeles​-de​-energia​-13/. Tolmasquim, M. T., T. de Barros Correia, N. Addas Porto, and W. Kruger. (2021). ‘Electricity market design and renewable energy auctions: The case of Brazil’. Energy Policy, 158, 112558. Winkler, J., M. Magosch, and M. Ragwitz. (2018). ‘Effectiveness and efficiency of auctions for supporting renewable electricity – What can we learn from recent experiences?’ Renewable Energy, 119, 473–489. https://doi​.org​/10​.1016​/j​.renene​.2017​.09​.071. Winkler, J., B. Woodman, A. Billerbeck, V. Anatolitis, and P. Kerres. (2022). ‘Auctions beyond electricity: How auctions can help to reform the energy system of the future’. AURES II Report D7.4. http://aures2project​.eu​/2022​/10​/05​/auctions​-beyond​-electricity/. Yalılı, M., R. Tiryaki, and M. Gözen. (2020). ‘Evolution of auction schemes for renewable energy in Turkey: An assessment on the results of different designs’. Energy Policy, 145, 111772. Zheng, L., V. Anatolitis, and J. Winkler. (2022). ‘Which support instruments can be used to promote green hydrogen? – Lessons learned from renewable electricity support schemes’. In: 18th International Conference on the European Energy Market (EEM).

18. The role of design elements in instrument mixes: the case of auctions and renewable portfolio standards in South Korea Tae-Hyeong Kwon and Pablo del Río

INTRODUCTION It is widely acknowledged that electricity from renewable energy sources (RES-E) provides various environmental benefits compared to electricity from fossil fuels (del Río and Gual, 2004). As shown in Chapter 16, countries worldwide have applied various deployment support policies to increase the market share of RES-E. These include administratively set support through feed-in tariffs (FIT) and renewable portfolio standards (RPS), with or without renewable energy certificates (RECs).1 Auctions have recently emerged as an attractive alternative and are now the fastest-growing and dominant instrument worldwide (del Río, 2017; International Renewable Energy Agency [IRENA], 2019; Grashof et al., 2020; Winkler et al., 2018). South Korea has relied on an RPS with RECs as the main instrument for supporting RES-E deployment. However, an instrument mix of auctions, RPS and FITs is now applied to mitigate the main shortcomings of RPS, that is, the volatility of revenues leading to substantial risks for investors. Numerous studies suggest that the success of instruments in general and auctions in particular depends on how those instruments are designed, i.e. on the choice of design elements (del Río, 2017; IRENA, 2015; Mora et al., 2017). Our starting point is that the design elements of instruments affect the interactions between the instruments within an instrument mix, and this is also the case for auctions. In other words, the design elements of auctions influence the synergies and conflicts between auctions and other RES-E support schemes. Although the analysis of the impact of the design elements of instruments on the functioning of those instruments has received some attention in the literature, this is not the case with their effects on instrument mixes. This issue has been under-researched, despite the widespread nature of instrument mixes, the importance of the choice of design elements in the successful functioning of instruments and their likely impact on the outcomes of instrument mixes. A notable exception is del Río and Cerdá (2017), who provide an analytical framework for discussing the impact of instruments and design features on the interactions between RES-E support and CO2 mitigation; they assess the comparative impact of different instruments and design features on the interactions between these instruments. This study thus builds on del Río and Cerdá (2017) and illustrates how the choice of design elements for one instrument may affect the outcome of the entire instrument mix. It examines the design elements of auctions in South Korea, assessing their effects on the instrument mix by identifying their impact on the interactions between instruments. Most contributions on combinations of policies (or policy mixes) are relatively recent but already quite abundant (see Rogge and Reichardt, 2016, for a review). Although many authors 420

The role of design elements in instrument mixes  421

define a policy mix as a combination of policy instruments which interact to influence a given outcome, whether it is R&D investments (Boekholt, 2010; Nauwelaers et al., 2009) or climate change and renewable energy targets (IEA, 2011; del Río, 2007), several authors argue that policy mixes encompass more than just a combination of policy instruments. For instance, Rogge and Reichardt (2015) and Flanagan Uyarra and Laranja (2011) argue that a policy mix also includes the processes by which instruments emerge and interact. Similarly, del Río (2014) observes that instruments are one of the components of policy mixes, along with policy framework conditions (e.g. goals, targets and stability). Rogge and Reichardt (2015, p. 7) define the policy mix as a combination of three building blocks: elements, processes and characteristics, which can be specified using different dimensions. Elements comprise the (i) policy strategy with its objectives and principal plans for achieving them and (ii) the instrument mix with its interacting policy instruments.

Therefore, an instrument mix is part of the overarching policy mix and probably its main component. Thus, the further analysis of instrument mixes is justified. Indeed, the studies on instrument mixes precede the broader policy mix contributions. Instrument mixes have been analysed for almost two decades in environmental and energy economics (IEA, 2011; Boots et al., 2001; Hindsberger et al., 2003; Jensen and Skytte, 2003; Johnstone, 2003; Morthorst, 2001; Sorrell et al., 2003),2 innovation policy (Nauwelaers et al., 2009; Flanagan et al., 2011) and economic policy more generally (Claeys, 2006). In environmental and energy economics, instrument mixes are defined as situations in which “several instead of one policy instruments are used to address a particular environmental problem” (Braathen, 2007, p. 186).3 In this context, the assessment of instrument mixes has focused on the conflicts between RES-E deployment support and CO2 mitigation instruments (e.g. an emissions trading scheme (ETS) or a carbon tax). This literature stream has shown that, in fact, such coexistence of instruments may lead to some problems, including redundancies and double counting (Sorrell and Sijm, 2005).4 It is logical to expect that the coexistence of different instruments would lead to conflicts, complementarities, synergies and interactions. Interactions between instruments broadly refer to the influence of one policy instrument being modified by the coexistence of other instruments (Nauwelaers et al., 2009). As argued by Bouma et al. (2019) and Greco et al. (2020), the interaction between different instruments may lead to positive, negative or neutral effects. Therefore, the outcome of the instrument mix strongly depends on these interactions. The interactions are clearly instrument- and context-dependent, which is why more empirical analyses of these instrument mixes and interactions are needed. These analyses may identify points of conflict between instruments (i.e. when they work against each other in a counterproductive manner), which changes in instruments or changes in the design elements within instruments may help alleviate. Indeed, as stressed by several extant studies, the instruments in an instrument mix need to be consistent, that is, they should reinforce rather than undermine each other in the pursuit of policy objectives. This is why strong instrument mix consistency is associated with positive interactions and weak instrument mix consistency is characterised by neutral interactions, while instrument mix inconsistency is captured by negative interactions (Rogge and Reichardt, 2016; Howlett and Rayner, 2013; del Río, 2010). As mentioned above, the instrument mix or policy mix literature usually analyses the interactions between instruments but does not consider the impact that the characteristics of the

422  Handbook on the economics of renewable energy

instruments (i.e. their design) have on those interactions. Indeed, del Río and Cerdá (2017) suggest that this is insufficient for analysing the interactions between the instruments in an instrument mix, because the instruments’ design (i.e. the choice of design elements) can significantly affect those interactions. Rogge and Reichardt (2015) consider two key attributes of policy instruments: type and design features. The latter correspond to “design elements”. Rogge and Reichardt (2015) argue that design features (or elements) can be differentiated into descriptive and abstract features, although the distinction between the two is somewhat artificial and some design features may belong to both categories (e.g. the level of support). Abstract design features refer to stringency, level of support, predictability, flexibility, differentiation and depth. Descriptive design features (or elements) refer to structural choices that constitute the content of an instrument, such as the actors or technologies eligible for support, the duration of support and other support conditions. They are regarded as the first step in identifying how a policy instrument performs when considering abstract design features. The design elements considered in this chapter are deemed the most relevant regarding their focus (renewable electricity auctions) and include elements in both categories (see below). The analyses of instrument mixes may have been too narrow or “bottom-up” and have neglected the broader picture and important elements. Thus, studies adopting a more comprehensive policy mix concept promise to significantly enhance our understanding of the link between policy and environmental technological change by going beyond the analysis of instrument interactions and including policy strategy, policy processes and characteristics (Rogge and Reichardt, 2015). However, it could also be argued that policy mix studies (but also the instrument mix contributions) have failed to assess the nuances of instruments and particularly the lowest granularity level (i.e. the design elements of instruments). Many studies in several areas have shown that the design elements of instruments critically influence the outcome of policies and instruments. This is very clear in the realm of renewable energy promotion and, particularly, in RES-E auctions. There is a strong consensus in the literature on RES-E auctions that the functioning of this instrument (and, thus, the functioning of the policy mix in which this instrument is embedded) is strongly dependent on the design elements chosen in the auction (del Río, 2017; IRENA, 2019, 2015; Mora et al., 2017; Kitzing et al., 2019). In summary, the interactions between instruments depend on the choice of design elements, but how the latter affect such interactions has been disregarded in the energy policy debate, particularly for RES-E auctions. Furthermore, although there is abundant literature on the design of renewable energy auctions, the analyses have often not considered that auctions are not implemented in a vacuum; they have not taken into account that the auctions are not isolated from the policies and instruments with which they interact. Therefore, this study attempts to uncover these two gaps in the literature. First, it empirically presents, using the case of the instrument mix in South Korea for the promotion of RESE, the importance of design elements in the interactions between instruments and provides insights on a largely neglected topic. Second, it contributes to the literature on the design of renewable energy auctions, broadening the assessment beyond the narrow focus on this instrument when it interacts with others in the instrument mix. The remainder of this chapter is structured as follows. The second section proposes an analytical and methodological framework (method and approach) for assessing the design elements of auctions and the instrument mix in South Korea. The third section describes the

The role of design elements in instrument mixes  423

auction design elements linking auctions to RPS. The fourth section assesses the impacts of auction design elements on REC prices, focusing on elements linking auctions to RPS, as well as on the instrument mix. The fifth section examines the actual trends of REC prices in South Korea based on the assessment in the fourth section. The sixth section concludes the chapter and presents policy implications.

ANALYTICAL AND METHODOLOGICAL FRAMEWORK: ASSESSING AUCTION DESIGN ELEMENTS AND THE INSTRUMENT MIX IN SOUTH KOREA Since this study analyses the impact of auction design elements on the interactions between instruments, an overview of these design elements is provided in the first subsection, followed by a description of the assessment criteria used to assess those design elements and their impact on the interactions in the second subsection, and the analytical and methodological approach in the third subsection. Design Elements in Auctions Auctions have been increasingly adopted for RES-E promotion because of their advantages in terms of efficiency and volume control (del Río, 2017; Lucas et al., 2017). Design elements matter for auctions to fulfil these expectations (IRENA, 2015; del Río and Linares, 2014). Table 18.1 describes the most relevant design elements in RES-E auctions.5 Table 18.1  Key design elements in RES-E auctions Design element

Description

Contract awarded (remuneration type)

• FIT/FIP/sliding premium • Contract period

Volume

Capacity (MW)/generation (MWh)

Periodicity

Regular rounds with a schedule/stand-alone auction

Qualification

• Specification of the offered project (technical requirement, documentation requirement and preliminary license) • Technical or financial capability of the bidding party • Economic guarantees (bid bonds)

Auction type

Sealed-bid auction/multi-round descending auction

Selection criteria

Price-only auction/multi-criteria auction

Pricing rules

• Pay-as-bid (PAB)/uniform pricing • Price ceiling/price floor

Diversity

Diversity with respect to technologies, locations, actors and size of installation

Realisation period

Deadline for building the project

Penalty

Forbid participation in successive auctions/reduce level of support/reduce the length of the support period/confiscation of bid bonds/penalty payments

424  Handbook on the economics of renewable energy

Assessment Criteria The assessment of auction outcomes considering other schemes, which partly depends on their design, has been conducted using different criteria. Effectiveness and efficiency are two of the most commonly used methods. Effectiveness refers to the extent to which an RES-E instrument encourages RES-E deployment (i.e. measured as either generation or capacity). In our case, which focuses on auctions for solar photovoltaic (PV) in South Korea, the effectiveness of the auction first refers to this technology. However, effectiveness with respect to the instrument mix, as in this case, should not only focus on one instrument (auctions) and one technology (solar PV) but should be a broader criterion for identifying the impact on other instruments as a result of the interactions and other renewable energy technologies. Therefore, the assessment of effectiveness should focus on this broader impact. Conversely, efficiency may refer to different aspects, including allocative efficiency, support cost efficiency and the minimisation of system generation costs (see del Río and Cerdá, 2014; Peñasco et al., 2019, for a full discussion). We focus on support cost efficiency, which involves achieving RES-E deployment at the lowest possible support costs. These costs are finally paid by consumers and, thus, governments are obviously highly concerned about their minimisation. In the case of auctions, lower levels of awarded bids mean higher levels of support cost efficiency. Considering the overall instrument mix, lower overall support costs for RES-E deployment (e.g. lower policy costs for the aggregation of all support schemes such as FITs, auctions and RPS with RECs) mean a higher support cost efficiency. Approach and Method Our starting point is that the choice of design elements in auctions influences auction outcome, as shown by several contributions (del Río, 2017; IRENA, 2015; Mora et al., 2017) and, thus, they influence policy mixes if auctions are part of them. In South Korea, auctions play a complementary role to the RPS, which is the main support scheme.6 Under RPS, RES-E suppliers receive a subsidy for their sales of RECs. The auction outcome is the sale of RECs.7 Since REC transactions in auctions are part of the REC transactions under the RPS programme, assessing the impact of auction design on the auction market and on other REC markets under the RPS scheme is important. Therefore, among the design elements of auctions, this study focuses on the design elements linking auctions to RPSs through their impact on REC markets. In particular, it assesses their impact on REC prices, which can be regarded as support costs. Similar to any other market goods, REC prices are determined by the demand and supply of RECs. While REC demand is equal to the REC allocation determined by the RPS target, REC supply is determined mainly by the expected revenues and costs for RES-E. The risk of future changes in expected revenues is also an important factor in determining the REC supply. Auction design elements thus affect REC prices in the auction market, with an impact on the REC transaction volume and the risk or revenues of the RES-E supply. Additionally, the changes in volumes or prices in the auction market also affect other REC markets. The assessment of the role of design elements in the interaction between instruments is performed relying on a theoretical comparative analysis of one design element with respect to its alternative, and the effect of the alternative design elements on the instrument mix is supported and illustrated with official data and trends. Figure 18.1 illustrates the research framework of this study.

The role of design elements in instrument mixes  425 DESIGN ELEMENTS Aucon design elements

INSTRUMENT MIX

AUCTIONS

FITs/FIPs

RPS (RECs) OTHER

OUTCOME - Aucons - Policy mix

Figure 18.1  Research framework flowchart

AUCTION DESIGN ELEMENTS LINKING AUCTIONS TO RPS IN SOUTH KOREA RPS in South Korea The main support scheme for RES-E in South Korea is the RPS. Under the RPS, 21 power supply companies whose capacities are over 500 MW have RES-E targets. The RES-E target increased from 2% in 2012 to 10% in 2022. A power supplier can fulfil its target by supplying electricity solely from RES-E or by obtaining RECs from the REC market. RECs were issued for all RES-E generation. One REC represents 1 MWh of the RES-E. RES-E suppliers can generate revenue by selling RECs and electricity. RPS poses a risk of rapid revenue changes for RES-E suppliers (Kwon, 2018). The revenues of RES-E suppliers are the sum of REC prices and the wholesale electricity price (system marginal price or SMP). Both REC prices and SMP are vulnerable to rapid changes in the RPS. The competitive feature of RPS/REC schemes is the main advantage of this instrument, whereas the higher risks of these schemes (due to volatility of revenues due to uncertainty on the evolution of the REC prices) is a well-known weakness (Ragwitz et al., 2007; del Río, 2005; IEA, 2008). However, is it possible to mitigate these risks without losing their competitive features? Setting ceilings and floors for REC prices is a potential solution. Another approach explored in this chapter involves combining instruments. The Instrument Mix in South Korea: Auctions, RPS and FITs In 2017, South Korea introduced long-term contract auctions with a sliding feed-in premium (sFIP). The key design elements of these auctions can be summarised as follows. The auction is part of the RPS scheme. The contracts awarded are for a 20-year supply of RECs by (one-sided) sFIP. Auction volume is capacity-based (MW) and is determined by the sum of the demand by power suppliers with an RPS obligation. There is a minimum requirement for public-owned power suppliers to buy RECs in the auction market. The auction was introduced

426  Handbook on the economics of renewable energy

Table 18.2  Summary of key design elements of auction in South Korea Auction design element

Auction design of South Korea

Contract awarded (remuneration type)

• Sales of RECs for 20 years • (One-sided) sliding FIP

Volume

• Capacity (MW) • Sum of the volume requested by power suppliers with RPS obligation • Minimum requirement for public-owned power suppliers

Qualification

Building permission for power supply

Selection criteria

Multicriteria (75–85 score of a total of 100 determined by price only)

Auction type

• Static sealed bid • Pay as bid (PAB)

Pricing rule

• Price ceiling • FIT level based on auction results

Periodicity

Twice a year

Diversity (size)

50% for small solar PV (less than 100 Kw) (50% of auction volume)

Diversity (technology)

Solar PV only

Realisation period

• 12 months (for more than 1 MW capacity) • 5–7 months (for less than 1 MW capacity)

Penalty

Suspension for 3 years

only for solar PV. There is also a 50% priority for small solar PV installations below 100 kW. The auctions are held biannually, in April and October, and are sealed-bid auctions with payas-bid. There is a ceiling price but not a floor price. In 2018, South Korea introduced a FIT scheme for only very small solar PVs ( Contract price, P(REC)=0

Contract Price

Transaction during the contract period (20years)

Figure 18.3  Long-term contract price of auctions in South Korea

430  Handbook on the economics of renewable energy

Table 18.4  Impact of auction remuneration types on REC markets Auction design (contract awarded/ remuneration types)

Impact on REC market

Impact on REC prices

• (One-sided) sFIP for a 20-year contract of REC sale

• Reduction of market risk for RES-E suppliers • Increase of RES-E supply

↓Price of RECs (fFIP) ↓↓↓Price of RECs (sFIP) ↓↓ Price of RECs (CfD)

investors (Noothout et al., 2016; Rathmann, 2011). These lower risks for RES-E investors and generators under an sFIP would lead to a more attractive scheme for these actors, greater participation (compared to fFIPs) and, thus, a higher RES-E supply, which would lead to a lower REC price. With a two-sided sFIPs (the so-called contract-for-differences [CfD]), a reduction in REC prices could be expected compared to fFIPs, given the lower risk and higher revenues for RES-E generators, which would tend to increase the RES-E supply. However, the impact of a CfD on the reduction of REC prices can be expected to be lower than for a one-sided sFIP, because a one-sided sFIP is more favourable for RES-E generators. Table 18.4 summarises this assessment. The lower risks for investors are evidently favourable for them, whereas a reduction in REC prices leads to lower support costs, which is favourable for consumers. However, the aforementioned risk reduction for investors does not necessarily involve a reduction of the risks for all actors. Indeed, there is a transference of risks between actors because, under a one-sided sFIP, the RES-E generators do not have to return to the electricity system the difference between the revenues from the SMP and sales of REC when these revenues are above the strike price in the auction. This comes at the expense of electricity consumers who pay the difference. Considering the assessment according to the two criteria (effectiveness and support cost efficiency), the analysis must be conducted for the overall instrument mix, that is, auction/ REC and other REC markets. Regarding the auction/REC, compared to an fFIP, an sFIP would lead to lower risks for investors and, thus, a greater level of solar PV generation, which would favour effectiveness. Concerning the impact on support costs, a reduction in the price of RECs can be expected under the sFIP compared to an fFIP. A lower price of RECs can also be expected in other markets (given the interactions between the different REC markets). This would lead to a lower incentive for RES-E investments and generation in other markets (i.e. lower effectiveness but higher support cost efficiency). When these effects are aggregated, a positive impact of the sFIP on the instrument combination can be expected for the support cost efficiency criterion compared to an fFIP. By contrast, the overall effect on effectiveness is unclear, but whether positive or negative, it would be insignificant as long as the RPS targets remain constant. Minimum Volume The auctioned volume in the South Korean auction is based on capacity (MW) and not generation (MWh). The total auction volume is the sum of the volumes requested by the power suppliers with an RPS obligation. There is a minimum requirement of auction volume for each public-owned power supplier. Therefore, publicly owned electricity suppliers must buy a certain number of RECs in the auction. The minimum requirement rule prevents auction

The role of design elements in instrument mixes  431

Table 18.5  Impact of auction volumes on REC markets Auction design (volume)

Impact on REC market

Impact on REC prices

• Capacity (MW) • Sum of volume requested by power suppliers with RPS obligation • Minimum requirement

• More auction volume, fewer REC transactions in REC spot market but greater incentive to participate in the auction

• Short-term: higher auction volume, higher auction price • Long-term: lower price of RECs (balancing effects)

Table 18.6  Impact of the FIT pricing rule on REC markets Auction design (pricing rule)

Impact on REC market

Impact on REC prices

FIT pricing rule

Linking auction price to FIT/RPS

Strengthening balancing effects

prices from falling sharply owing to an auction volume that is too small. However, a higher auction volume also leads to a decline in REC demand in the REC spot market over 20 years, which will lead to a reduction in REC spot prices. Since the auction price and REC spot price cannot move far away from each other, there will be balancing effects between those prices in the long run (Table 18.5). The existence of a minimum auction volume implies a minimum demand for RECs which, compared to the absence of a minimum volume, would prevent a sharp reduction in REC prices. However, this minimum volume would reduce the demand in the spot market, which would have an uncertain effect on the price of RECs. Overall, the impact of minimum volumes on the instrument mix is unclear when considering both assessment criteria. This is because the higher effectiveness in the auction/REC (given the minimum REC price level ensured by the existence of a minimum demand) would be offset by the lower effectiveness in the other REC markets due to the lower REC demand in the REC spot market. Conversely, the higher REC price in the auction/REC would be compensated by a lower price in the other REC markets, which would also lead to a nonconclusive outcome regarding support cost efficiency. FIT Pricing Rule The auction price is the FIT for small solar PVs (less than 100 kW) for the following year in South Korea. Since the FIT price level affects the REC spot market, the link between the auction price and REC spot price is strengthened by the FIT pricing rule. For example, if the auction price increases, the FIT price will increase according to the pricing rule, which helps attract more small solar PV suppliers to the market. This will lead to a decrease in the REC spot price and REC auction prices (in the long term). Therefore, the FIT pricing rule strengthens the balancing effects among the various REC prices. Table 18.6 summarises this assessment. Therefore, a lower FIT level compared to the FIT level that is not tied to the outcome of the auction can be expected. This is because the asymmetric information problem in

432  Handbook on the economics of renewable energy

administratively set remuneration would probably lead to a higher FIT level. In contrast, a lower FIT indicates a lower level of effectiveness but also lower support costs (i.e. higher support cost efficiency). Schedule Auctions are held in April and October annually. This periodicity of auctions ex-ante helps reduce the risks for RES-E generators and encourages the participation of RES-E generators in the market. The greater participation and higher RES-E generation would increase solar PV generation and the supply of RECs, thus leading to lower REC prices. The values can be expected to be positive for both the effectiveness and support cost efficiency criteria regarding auction/REC. A lower price of RECs can also be expected in other markets (given the interactions between the different REC markets). This leads to a lower incentive for RES-E investments and generation in other markets (i.e. lower effectiveness but higher support cost efficiency). When these effects are aggregated, the positive impact of an auction schedule on the instrument combination can be expected for the support cost efficiency criterion compared to its absence. By contrast, the overall effect on effectiveness is unclear but probably insignificant under a constant RPS target. This assessment is summarised in Table 18.7. Frequency A critical issue when conducting auctions is setting the frequency of rounds, that is, the number of annual rounds. This is a difficult task, and there is a risk of setting either too many or too few rounds. Many rounds have the advantage that they can allow bidders to recover (part of) their sunk costs if the same material prequalification (e.g. administrative permits, connection permits) can be used in different rounds. However, they could lead to a “thin market” problem in one specific round, if there is insufficient competition in that round, that is, if the volume of bids submitted is small compared to the volume auctioned and awarded in that round (see, e.g. the example of German PV auctions in IRENA, 2019). By contrast, too few rounds may lead to large volumes being auctioned in one round. This would also result in competition problems. Furthermore, less frequent auctions may increase the time cost of RES-E generators. The “optimal” number of rounds depends on the technology and situation of the market (del Río, 2017). However, it is generally recommended to hold at least one auction annually. Therefore, the two rounds in the South Korean auctions seem to provide a good trade-off between the frequency of auctions and minimising the sum of costs, that is, costs due to market thinness and time cost (Table 18.8). Two auction rounds annually also seem appropriate, considering five months to one year as the realisation period after contracts.

Table 18.7  Impact of a schedule of auctions on REC markets Auction design (schedule)

Impact on REC market

Impact on REC prices

Schedule of auctions

• Reducing uncertainty in the market: increase in the participation of RES-E generators and generation

Lower price of RECs gradually

The role of design elements in instrument mixes  433

Table 18.8  Impact of auction periodicity on REC markets Auction design (periodicity)

Impact on REC market

Impact on REC prices

Periodicity: twice annually

• Reducing uncertainty in the market: increase in RES-E generators • Increasing frequency: thin market/decrease in time cost • Decreasing frequency: thick market/increase of time cost

Lower REC prices (under optimal frequency)

Table 18.9  Impact of auction design for size diversity on REC markets Auction design (diversity: size)

Impact on REC market

Impact on REC prices

Size (50% priority for small solar PV)

Increase of small solar PV generators

Short-term: price up of auction Long-term: lower REC price

Compared to a low frequency of rounds, an optimal number of rounds reduces the costs for RES-E investors, increases participation in the auction and tends to reduce REC prices. It also discourages underbidding since some bidders are desperate to be awarded when there is only one round at irregular intervals. They believe that this will be “their last opportunity” to be awarded, which may lead to too low bids. Underbidding leads to underbuilding and, thus, low effectiveness (del Río, 2017). However, a lower REC price would reduce the incentive for RES-E investments and generation in other markets. Therefore, an optimal frequency is a good choice from a support cost efficiency perspective. The overall impact on effectiveness is unclear but probably insignificant under a constant RPS target. Diversity (Size) There is a 50% priority for small solar PVs (