Energy Research & Social Science 92 (2022) 102775 Contents lists available at ScienceDirect Energy Research & Social Science journal homepage: www.elsevier.com/locate/erss Original research article Why energy models should integrate social and environmental factors: Assessing user needs, omission impacts, and real-word accuracy in the European Union Diana Süsser a, Nick Martin b, Vassilis Stavrakas c, Hannes Gaschnig d, Laura Talens-Peiró b, Alexandros Flamos c, Cristina Madrid-López b, *, Johan Lilliestam a, e a Institute for Advanced Sustainability Studies, Berliner Straße 130, 14467 Potsdam, Germany b SosteniPrA Research Group, Institut de Ciència i Tecnologia Ambientals, Universitat Autònoma de Barcelona (ICTA-UAB), Z Building, Campus UAB, 08193 Barcelona, Spain c Τechnoeconomics of Energy Systems laboratory (TEESlab), Department of Industrial Management and Technology, University of Piraeus, Karaoli & Dimitriou 80, Piraeus 18534, Greece d acatech - National Academy of Science and Engineering, Karolinenplatz 4, 80333 München, Germany e University of Potsdam, Faculty of Economics and Social Sciences, August-Bebel-Straße 89, 14482 Potsdam, Germany A R T I C L E I N F O A B S T R A C T Keywords: Energy models are used to inform and support decisions within the transition to climate neutrality. In recent Energy modelling years, such models have been criticised for being overly techno-centred and ignoring environmental and social Energy planning factors of the energy transition. Here, we explore and illustrate the impact of ignoring such factors by comparing Energy policy model results to model user needs and real-world observations. We firstly identify concrete user needs for better Ecological crisis representation of environmental and social factors in energy modelling via interviews, a survey and a workshop. Social acceptance Environmental impacts Secondly, we explore and illustrate the effects of omitting non-techno-economic factors in modelling by con­ trasting policy-targeted scenarios with reality in four EU case study examples. We show that by neglecting environmental and social factors, models risk generating overly optimistic and potentially misleading results, for example by suggesting transition speeds far exceeding any speeds observed, or pathways facing hard-to- overcome resource constraints. As such, modelled energy transition pathways that ignore such factors may be neither desirable nor feasible from an environmental and social perspective, and scenarios may be irrelevant in practice. Finally, we discuss a sample of recent energy modelling innovations and call for continued and increased efforts for improved approaches that better represent environmental and social factors in energy modelling and increase the relevance of energy models for informing policymaking. 1. Introduction optimisation and economic costs or benefits of the energy transition and do not fully address multiple dimensions of truly sustainable pathways, The European Union (EU) has set the goal of transitioning to a including regional environmental impacts, material requirements of modern, resource-efficient, and competitive European economy, with energy technologies, diverging normative views or citizen preferences. the overarching objective of climate neutrality by 2050 [1]. The energy This imbalance, where energy policy is determined at the expense of transition is crucial to this plan and is a cross-societal process, including factors outside the techno-economic realm, is also reflected in current both socio-technical and socio-ecological drivers and constraints that energy modelling practices. underlie the required system changes [2]. EU energy policy strategies Most energy models used to inform the energy transition ignore under the “European Green Deal” emphasise the need to develop energy factors other than techno-economic ones, generally seeking cost-optimal systems that provide secure, affordable and clean energy, reduce envi­ futures. They rarely consider environmental factors beyond greenhouse ronmental impacts, and enable citizens to participate and benefit [3,4]. gas (GHG) emissions. For example, integrated assessment models (IAMs) Nevertheless, most visions and policy goals concern the technological typically only include simplified emission and land-use assumptions [5]. * Corresponding author. E-mail address: cristina.madrid@uab.cat (C. Madrid-López). https://doi.org/10.1016/j.erss.2022.102775 Received 27 April 2022; Received in revised form 25 July 2022; Accepted 16 August 2022 Available online 20 September 2022 2214-6296/© 2022 Published by Elsevier Ltd. D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Modellers often entirely ignore social aspects, or only consider them as materials, production, transportation, and installation of components, an exogenous narrative do be discussed “on top” of techno-economic and the ongoing maintenance and eventual decommissioning of plants – findings, as a lens through which techno-economic scenarios can be are often not accounted for and remain “hidden” [36]. discussed [6]. Nevertheless, there is growing recognition that environ­ The need for raw materials is another issue that has gained mental and social factors must be included in models [7,8]. As with increasing public and political attention as the ongoing production of present energy systems, future decarbonised energy systems will face many sustainable energy technologies – particularly wind turbines, solar environmental constraints such as raw material or water availability photovoltaic (PV) cells, and lithium-ion batteries – require supplies of [9,10]. Presently, public opposition against energy infrastructure pro­ critical raw materials (CRMs) [37]. For example, Europe is 100 % import jects is halting transition progress in Europe and across the world [11]. reliant on borates, lithium, and graphite for EV batteries, silicon metal Ignoring such factors risks producing mathematically elegant but for photovoltaic panels, niobium for permanent magnets in wind tur­ politically irrelevant scenario results. At the same time, modellers are bines, and a mix of diverse rare earth elements for EV batteries and bound by model and computational capacities [12] and will only include permanent magnets [38]. China remains the dominant provider of factors that are easily quantifiable or do not challenge the disciplinary processed materials and components [37,38]. Dependency on scarce barriers of their respective modelling frameworks [13]. raw material often leads to geopolitical clashes, “carbon leakage” These challenges result in a gap between the information provided [25,40,41], externalisation of impacts [42] or environmental dumping by energy models and the information needed by those who use the [43]. Greater adoption of material reuse and recycling could help to model results. Scholars have identified gaps related to the modelling of alleviate such pressures [44] and, therefore, strengthening the circular behavioural and lifestyle changes [14], specific policy challenges [15] economy has become a key strategy within the EU [45]. and modelling of political or societal paradigm shifts [16]. Neglecting Quantifying the impacts of energy infrastructure on land, water, and these factors may result in energy policy goals or implementation stra­ biodiversity is also gaining attention, particularly within the growing tegies that conflict with environmental policy [17], or undermine social literature surrounding the water-energy-food nexus [46]. For example, goals unknowingly [18]. As a result, oversimplified models could fail to impacts relating to land occupation have been identified for wind and inform policymakers about the multiple dimensions crucial for a sus­ solar installations [47,48], land-use impacts and water overexploitation tainable energy transition. are often linked to bioenergy [49], and biodiversity issues can be linked Here, we investigate concrete needs for better representation of to hydropower, marine and geothermal energy [50]. Although some environmental and social factors in energy modelling and explore the studies have investigated land-use for solar farms [51], onshore wind implications of current model shortcomings. In a recent study [19], it turbine siting remains the most prominent example of land-use conflicts was shown that both model users and modellers see a need for improved regarding renewable energy technologies [52,53]. Finally, water re­ representation of social and environmental aspects in modelling. Here, quirements for different energy production options is gaining attention we advance the analysis by analyzing more deeply what concrete as southern and more arid countries seek to adopt cleaner technologies environmental and social factors users consider most important for and general awareness of water availability issues grows [54]. better representation in energy models and by examining why these Nevertheless, despite the importance of these aspects, most of the factors are important and illustrating their importance. We adopt a two- models used to inform energy policy are limited in their consideration of fold approach: first, we empirically identify and rank specific user environmental factors. First, most accounting methods only consider modelling needs for environmental and social aspects through in­ direct emissions, and the indirect emissions and other impacts embodied terviews, a survey and an online workshop with different model users. within energy processes. Second, CRMs are generally not considered in Second, we investigate and illustrate the magnitude of problems arising any detail, particularly not in the large-scale models being used to through the omission of central social and environmental factors inform overarching climate policy. Third, although land availability is­ through real-world case studies. We show that the impact of omitted sues continue to be an issue in energy planning processes [55–59], it is social and environmental factors could be so large as to render results generally only modelled as a constraining factor for technical potentials unfeasible and hence irrelevant, highlighting the necessity to consider and societal or political preferences for present or future land use are non-technoeconomic factors as integral parts of energy models. largely ignored. 2. Background on environmental and social factors and energy 2.2. Social drivers and barriers to the energy transition modelling While environmental aspects are considered constraining factors to A wealth of literature exists that addresses the different environ­ most transition options, social aspects can both accelerate or impede mental and social factors that can drive or hinder energy transitions. them [11]. Although the transition to renewable energy enjoys high Many studies investigate environmental impacts of renewable electricity public approval levels within the EU [60], concrete projects often face production [20,21], storage [22], electric vehicles (EVs) [23], material considerable opposition [11,61]. Issues typically relate to the increasing dependency [24], or emissions [25]. Other authors investigate social number of renewable energy plants and associated transmission infra­ issues of energy transitions, such as behaviour and lifestyle [26–28], structure, conflicts arising from place attachment [62], planning and public acceptance [29,30] and ownership [31,32]. Some studies also siting issues [63], visual and aesthetic impacts [64], land-use conflicts integrate both perspectives, by addressing environmental justice in en­ [65], biodiversity loss [66–68], and noise, or health concerns [69]. ergy and climate policy, for example [33,34]. Below, we describe and Accordingly, the social acceptance of strategies and projects is gaining discuss the current state of the literature regarding environmental and importance as the transition accelerates towards 2030 and 2050 targets. social factors in the energy transition and in modelling. This includes not only acceptance of technologies, but also of new end- use services or practices and lifestyles or cultural meanings of energy 2.1. Environmental implications of the energy transition [70]. The effort to increase awareness and acceptance accompanies calls for comprehensive citizen participation and ownership [71–73] and The reduction of carbon dioxide (CO2) and other GHG emissions research continues about the ways that local populations make choices needed for mitigating climate change dominates the debate about about consumption and investments [74], and how social acceptance is environmental impacts in the energy sector [35], although policy de­ formed. However, this knowledge is yet to be widely integrated into cisions and models generally only depict the direct emissions during the energy models. final stages of energy production. Indirect emissions related to other The energy transition has given rise to a new generation of agents stages of production life cycles – e.g., those related to extraction of raw who take on the role of active producers, distributors, consumers, and 2 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 sellers of renewable electricity, the so-called “prosumers.” Citizens may In any case, while current energy models rarely represent these social become owners, eventually consuming their own electricity, or become factors, different model types do offer some capabilities [6]. For part of community energy projects [75], potentially bringing local example, more nuanced bottom-up modelling approaches, like agent- benefits, such as employment and increasing project acceptance [72]. based models (ABMs), can address social barriers for solar PV adop­ Still, the advantages and mechanisms that allow citizens to participate in tion [32] or peer-to-peer energy trading in local communities [87], transition processes are also generally excluded from energy models. while demand models can address drivers and patterns of household Furthermore, many researchers have studied how norms, practices energy consumption [28]. However, significant modelling gaps exist, and culture shape energy behaviour [76] and how consumer behaviour especially when dealing with transition dynamics, e.g., speed of trans­ and lifestyle affect climate change mitigation [77]. Despite the high formations and path dependencies [7], and socio-technical systems that environmental awareness among citizens in industrialised countries, captures agent heterogeneity, e.g., zero-energy communities [88]. behavioural changes and sufficiency-based lifestyles are still relatively uncommon for reasons such as lack of awareness, comfort, fear of loss, or 3. Methods exclusion [78]. In contrast, behavioural change is often seen not as a welfare loss but as a gain in wellbeing and satisfaction [79], as beneficial To highlight the relevance of including social and environmental lifestyle innovation [80], and the “holy grail” of sustainability [81], aspects in energy modelling, we build upon a previous, related study particularly outside mainstream economics. Nevertheless, energy suffi­ [89], where it was shown that, in general, environmental and social ciency remains a marginal strategy in energy policy documents aspects are relevant to modellers and users of model results. Here, compared to energy efficiency and renewable energy sources [82]. working with the same data, we further investigate what environmental Many models, however, assume that lifestyle changes are happening and and social factors of the energy transition are relevant for inclusion in demand-side measures have gained increasing interest to initiate con­ energy models and perform a stakeholder-based ranking of importance. sumer behaviour. We then investigate case studies where models have ignored these As behaviour is strongly guided by routines, public policy plays a central environmental and social factors and illustrate their importance central role in adapting behaviour, including modifying consumption by comparing them to real world developments. As depicted in Fig. 1, we and investment choices [83] and in municipal renewable energy completed both empirical and desk research. The empirical research was deployment [84]. For example, some EU member states, including implemented within the context of the EU-funded Horizon 2020 project Germany, Spain and Denmark, implemented feed-in laws in the 1990s, “Sustainable Energy Transitions Laboratory (SENTINEL).” As this work thus supporting the early adoption of renewable energy technologies by coincided with the COVID-19 pandemic, we conducted all stakeholder individuals and municipalities [85,86]. engagement activities in an online format, as was common practice in Fig. 1. Two-step approach employed within the study, consisting of empirical and desk research components. 3 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 the EU energy research community at the time [90]. 4. Results 3.1. User needs: identification and ranking 4.1. User needs on environmental and social aspects of the energy transition We conducted 32 interviews in five jurisdictions – the EU, Germany, Greece, Poland, and Sweden. This included four different stakeholder Our findings suggest that model users want better integration of groups that participate in modelling-informed energy policymaking in different environmental and social factors of the energy transition in Europe: scientists, non-governmental organisations (NGOs), energy in­ models, particularly with respect to raw material demand/availability dustry experts and policymakers. The interviews were guided by a semi- and natural impacts (environmental) and social acceptance, consumer structured guideline, were conducted in English, or in the national behaviour and policy dynamics (social). Our results show that users language of the jurisdiction, and all interviews were transcribed and prefer the explicit integration of social and environmental aspects over anonymised after being recorded. More information is provided in the further improvements of techno-economic aspects: The workshop par­ Appendix (Table A1). ticipants ranked “Impact on the environment and natural resources” Building on the interview findings, we performed a Europe-wide highest, followed by policy impacts, social impacts and costs (see Ap­ survey to obtain deeper insights about which social and environ­ pendix Fig. A1).The high-level results from the survey have been re­ mental factors are important from a larger stakeholder sample. The ported in [19,92]. Here, we present in more detail the environmental survey was designed as a semi-quantitative online questionnaire and and social aspects model users see as particularly important, how they contained different question formats, from single and multiple choice to ranked those aspects and reasons provided for their relevance. Likert-like scales and free-text boxes, depending on the variables to be addressed. The survey was distributed among national, European and 4.1.1. Environmental aspects international organisations, to representatives from politics, civil soci­ Our results identify raw material use and material circularity as ety, business/industry and research, via private and public online central model user concerns. More than half of the survey respondents channels. We asked questions regarding what factors should be better stated that they would like to see raw material demand integrated into represented by models and asked specific follow-up questions regarding energy models, followed by GHG emissions, air pollution, water usage environmental and social aspects. In all, we received a total of 90 and loss of biodiversity (see Fig. 2), although relevance varies strongly completed questionnaires. Further information on the survey can be by user group. Energy industry and researchers tended to prioritise GHG found in the Appendix (Table A1) and the questionnaire and anony­ emissions and air pollution, whereas NGOs and policymakers expressed mised aggregated data are available at Zenodo [91]. greater concern about raw materials, water issues and biodiversity. One Finally, we discussed and ranked the environmental and social fac­ NGO representative underlined this by saying: “Also, the whole envi­ tors identified in a workshop with stakeholders from different EU ronmental aspects, like the biodiversity aspect of wind energy... we can't member states. The workshop allowed us to discuss specific user needs achieve 100% renewables without having hundreds of gigawatts of offshore in more detail and collect more data on the different aspects. We held wind. That is going to be crucial, but you also have to do it in a sustainable one breakout session on modelling of environmental aspects and one on way” (EU_NGO#2).1 Another interviewee added that “It is a question of modelling of social aspects. For the social aspects, we integrated the resource efficiency. The resources to reduce climate gases, but also that we ranking results from two live polls in two breakout sessions. For envi­ need to use the resources we are having as efficient as possible – also if it's ronmental aspects, attendants to two breakout sessions discussed and waste, we are using” (Sweden_science#4). agreed a ranking. Accordingly, only integrated results are shown here. In the interviews and at the workshop, methods to capture the full Furthermore, because only 25 stakeholders participated in the work­ life cycle of energy technologies and infrastructures, and not only direct shop, we did not distinguish between specific stakeholder groups. impacts, and the degree of externalisation of impacts that can be Further information on the workshop can be found in the Appendix observed in the literature, were further central concerns. The relative (Table A1). importance of environmental aspects was also explored within the workshop. When asked to rank factors, participants identified four as­ 3.2. Case studies on omitting environmental and social factors in energy pects of particular importance: (E1) Raw materials, (E2) Biodiversity, modelling land use, and water use, (E3) Life-cycle perspective, and (E4) GHG emissions beyond CO2 (Table 1). The inclusion of environmental aspects For the top-ranked user needs (Section 4.1), we identified and goes beyond the need to protect our ecosystems and natural resources. selected specific cases where energy system models have neglected Indeed, the main reasons argued for the need of including environ­ environmental and social factors (Fig. 1). These cases illustrate the type mental aspects were i) to support decision making processes, ii) to and magnitude of problems on the relevance of model-informed poli­ enable links to other models, policies, and strategies, and iii) to facilitate cymaking that may arise when models ignore these factors. We subse­ citizen empowerment and stakeholder engagement. quently selected four case studies in which model output and observed development are strongly misaligned because a critical social or envi­ 4.1.2. Social aspects ronmental factor was ignored by the model, for different European A high demand was also found for the better representation of social contexts. For each case, we conducted a document analysis of modelling aspects in energy models. When asked to select aspects that require applications and compared these findings with real-world developments better integration, participants nominated three aspects most often: co- and policy targets. Our goal was not to demonstrate that models fail to benefits of prosumerism and community energy, social drivers and predict the future, as this is not their aim; most models are used to barriers of innovation diffusion, and dynamics of social acceptance and explore possible (simplified) systems and investigate options and sen­ individual attitudes (Fig. 3). Here again, different stakeholder groups sitivities. Rather, we illustrate the importance of environmental and differ in choice frequency. Social acceptance and individual attitudes social concerns within models so that deeper and more robust un­ were more often chosen by NGOs, whereas policymakers raised concern derstandings of the mechanisms of transition pathways and more policy- about the impacts of social issues on politics and policies more often relevant model advice can be obtained in the future. than other users. Both NGOs and researchers agreed that benefits of 1 We apply the interviewee referencing (jurisdiction_stakeholder group_ interviewee number). 4 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Fig. 2. Key environmental aspects identified by the user needs survey (choice frequency; up to three answers possible, voluntary question). Responses were obtained for the following question: “You stated that environmental, or resource-relevant issues should receive more attention by energy models. What environmental factors would you like to see integrated into energy models more in the future?”, N = 47. also revealed the relevance of a better integration of policies in energy Table 1 modelling, going beyond CO2 prices as the only policy measure for Ranking of environmental factors to be included in energy models. prioritisation. During the discussion, participants expressed the need to Ranking of environmental factors understand the science and to compare it with ongoing policy processes, 1 Raw materials and to understand how policy changes can trigger behavioural changes. 2 Biodiversity, land use, and water use Many stakeholders also raised questions in connection to the choice of 3 Life-cycle perspective policy instruments for reaching targets. One interviewee asked: “And in 4 GHG emissions beyond CO2 the area of policy instruments, how you can move faster with the climate action? What instruments do we need?” (Sweden_scientist#4). individual and community participation should receive more attention. In the workshop, we asked stakeholders to rank different social as­ One interviewee highlighted the interlinkage between these different pects in two breakout sessions and two subsequent rounds of live poll­ factors: “…we have connections to social acceptability, because if we go into ing, finding the most important aspects to be (S1) Social acceptance/ a more decentralised approach, we can create more value for the regions, or opposition, (S2) Individual and community participation, (S3) Con­ for all European places, where you have your own creation of energy and you sumer behaviour and lifestyle, and (S4) Policy dynamics (Table 2). One have your own value chains. You have local jobs, local economy, and then, interviewee confirmed the limitations of current energy models by local acceptance” (EU_NGO#1). stating: “It can be in terms of social acceptance, it can be in terms of job In addition to social aspects, discussions in the stakeholder workshop creation, it can be in terms of socio-economic impacts that are not all factored Fig. 3. Key social aspects identified by user needs survey (choice frequency; up to three answers possible, voluntary question). Responses were obtained for the following question: “You stated that social aspects should receive more attention by models. What social aspects would you like to see integrated into energy models more in the future?” (voluntary, multiple choices, up to three an­ swers), N = 49, Explanation of terms: “Social acceptance” refers to the willingness of people to accept the installa­ tion of energy-related infrastructure, usually near them. “Optimisation of acceptance” refer to the aim of making resistance to installation as low as possible. “Social storylines and scenarios” refer to scenarios that include qualitative storylines, describing also societal de­ velopments and interactions and interdependencies be­ tween actors, technologies, and policy interventions in the context of the energy transition. 5 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Table 2 proposed projects may be delayed or not built at all. For example, the Ranking of social factors to be included in energy models. main scenario of the German Advisory Council for the Environment Ranking of environmental factors projected 42 gigawatts (GW) of interconnection between Germany and Denmark, and 48 GW crossing the Skagerrak to the hydropower stations 1 Social acceptance/opposition 2 Individual and community participation in Norway; in their “Supergrid” scenario, these interconnectors are 53 3 Consumer behaviour and lifestyle GW and 116 GW, respectively [94]. In 2020, the German-Danish 4 Policy dynamics interconnection capacity was 1.7 GW northward, with an ongoing expansion project to 2.5 GW [95]. The Danish mainland is just above 50 km wide at the narrowest place, suggesting that, if lines are land-based, in the model that is being run” (EU_industry#2). One policymaker added these scenarios imply on average 1–2.5 GW of transit powerline per that “[t]he improved simulation of ‘real-world’ decision-making and cross-section kilometre in Denmark. This casts great doubt on the behavioural aspects is always welcome and offer robust results in the quan­ feasibility – especially the social and political feasibility – of such a plan: titative analysis” (Greece_policymaker#1). The main reasons argued for Will Denmark accept such enormous capacity lines, especially if they are the need of including social aspects were i) to better understand people's merely passing their country with no immediate benefit to them? decision-making processes and criteria, and lifestyle choices, ii) to Indeed, opposition from transit countries has been problematic in past enable citizen and community participation in the energy transition, and projects, including the Desertec plan to import solar power from iii) to understand the (distributional) effects of different policy Morocco to Germany [96]. measures. Most political visions and plans are model-supported. For example, the European Commission (EC) use the EU Reference Scenario as a central 4.2. Case studies on the importance of environmental and social factors basis for their decisions [97]. Furthermore, many models, such as PRIMES, assume that the infrastructure plans within their simulations To illustrate the potential effects of omitting the top-ranked envi­ are completed as intended [97]. ENTSO-E's Ten-Year Network Devel­ ronmental and social factors from energy modelling (Section 4.1), we opment Plan (TYNPD) 2020 expects that over 300 transmission projects undertook four case studies. We describe these case studies and the of some 45,000 km will be commissioned by 2040 [98], with about 50 % omitted user needs discussed in each in Table 3. Note that we do not of projects expected to be operational by 2021–2025 (Fig. 4). However, explicitly illustrate point E3 (life-cycle perspective) because it is implicit if such plans do not materialise, models using this assumption clearly in the materials issue and point E4 (GHG emissions beyond CO2) because produce less meaningful results. reducing emissions is understood to be the key motivation of the energy In reality, implementation has been slow, with only 40 % of projects transition and a key variable in most energy models. For each case, we on or ahead of schedule, and all others delayed or altered in various present published model scenarios – or, in some cases, the lack of suit­ ways (Fig. 5). In 2020, 65 TYNDP transmission projects (17 %) were able outputs – alongside real-world situations to highlight mismatches reported as delayed, and this only includes early projects (2021–2025) between model results, real-world developments and policy targets. as later projects have not yet entered stages in which delays can occur [99]. This has not changed significantly over time: in 2012, a third of 4.2.1. The EU electricity grid plan without people and nature projects were reported as being delayed due to “social resistance and Transmission grid expansion is a key pathway for integrating fluc­ longer than initially expected permitting procedures” [100]. Pall et al. [101] tuating renewable supplies into power systems. Many modelling studies investigated the causes of deployment delays in international power show that new transmission lines must be built for a least-cost electricity transmission projects and found that local public resistance and political system in the EU. Rodríguez et al. [93] quantified the benefit of power interventions (strikes/blockades) are the main reason, while other transmission between countries to support almost 100 % renewable research found that public opposition is the most important delay factor power, finding a cost-minimum for a grid five times as large as today's. in national projects [102]. Underlying causes could be environmental Similarly, Tröndle et al. [29] found that the cheapest, continent-wide, concerns related to new grid infrastructure, as new expansions in power fully renewable electricity supply would require twice the present lines have the potential to harm local environments, and impact biodi­ transmission grid. However, they also show that if the transmission grid versity during both the construction and operation phases [103]. is used for the continental-scale balancing of net self-sufficient regional In sum, there is a large gap between what scientific and advice- supplies, much less transmission capacity – roughly the size of today's oriented models project and what is observed on the ground in trans­ transmission system, but with twice the cross-border capacities – would mission projects: not only are network plans much smaller than the vast- be required. Most cost-optimised renewable power scenarios critically scale expansion that cost-optimising models find beneficial, but actual hinge on the realisation and feasibility of grid expansion. progress is typically much slower than models optimise/simulate. This Beyond grid expansion, such scenarios often envisage large concen­ means that system models risk generating meaningless findings, high­ trations of generation and transmission at specific locations [29]. lighting the problem of ignoring social factors in technical models. Therefore, local acceptance is an essential factor; if citizens in these key places do not accept the plans, the scenario becomes irrelevant as the 4.2.2. An environmental dilemma for electric vehicles in the EU? The EU aims to reduce CO2 emissions from cars by 55 % (compared Table 3 to 1990) by 2030 and proposes to ban sales of fossil-fuelled cars by 2035 Identified case studies for demonstrating the importance of integrating envi­ [104]. Electrification of transport plays an important role in reaching ronmental and social aspects into energy modelling. net-zero emissions by 2050 [105]. A variety of energy models have Case study User needs explored future EV penetration rates and project EV use to increase The EU electricity grid plan without people (E2) land use dramatically in coming decades. For example, the International Energy and nature (S1) social acceptance / opposition Agency (IEA) Mobility Model projects 16 million electric cars in their An environmental dilemma for electric (E1) raw materials Stated policies scenario in the EU by 2030 and 33 million in their Sus­ vehicles in the EU? (S3) consumer behaviour tainability development scenario [106]. Statharas et al. [107] quantita­ Headwind for onshore wind power in (E2) biodiversity and land use Germany (S1) social acceptance / opposition tively assessed the impacts of factors that drive market penetration of Domestic investment behaviour for small-scale (S2) Individual and community electric cars in the EU, using the PRIMES-TREMOVE model, and project PV in Greece participation under the most optimistic scenario, that 18 % of the total car fleet in the (S3) consumer behaviour EU will be electric by 2030. (S4) policy dynamics Although energy models include detailed analyses of the transport 6 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Fig. 4. Transmission project timeline in the 2020 ENTSO-E Ten-Year Network Development Plan. Data source: ENTSO-E [99]. two-fold challenge: not only are many needed materials scarce in gen­ eral, but almost all of them are not produced domestically, making the European EV strategy vulnerable to supply risks. Bobba et al. [38] made material demand projections for lithium-ion battery production for three future e-mobility pathways, showing that even the lowest deployment pathway for batteries alone requires several times the present total EU consumption of lithium, cobalt and graphite (Table 4). This suggests that serious limitations may occur unless the EU can drastically increase its supplies of the three CRMs considered. Even if the EU imports all its EV batteries [117], global resources and pro­ duction remain limited, showing that there is a very real threat to the accelerated uptake of EVs, both in Europe and globally [118]. This raises the question of how feasible such projections are if they neglect material Fig. 5. Progress of all transmission investments since TYNDP 2018, n = 321 constraints, or whether sufficiency strategies for avoiding mobility, or projects. “Ahead of time”: expected commissioning date is earlier than antici­ shifting to other modes of transport, do not need to be pushed much pated in the previous TYNDP. “Rescheduled”: commissioning has been post­ more in the social and political debate. poned due to a voluntary decision. “Delayed”: expected commissioning date is later due to delay (unvoluntary). “On time”: no change compared to previous TYNDP. “New investments”: new in the TYNDP 2020 (in comparison to 2018). 4.2.3. Headwind for onshore wind power in Germany Data source: ENTSO-E [99]. Wind power substantially contributes to the power mix in Germany. In 2020, around 18 % of gross electricity production came from onshore and about 5 % from offshore plants [119]. In December 2021, the sector and EV numbers, very few consider raw material requirements as German onshore wind power capacity was 56 GW [120] and, as the a potentially constraining factor in their calculations. Yet, such factors energy transition progresses, it will likely become the most important may prove critical. For example, Xu et al. [108] developed a material electricity source [121]. The Renewable Energy Act 2021 aims for 71 flow analysis showing that global EV battery demand would increase GW capacity by 2030 [122], while the government's long-term climate key minerals consumption by a factor of 20–30 by 2050. While progress scenarios foresee 80 GW [123]. These targets equal an annual average has been made to develop methods for assessing material requirements expansion of 1.7–2.7 GW/year. [109] and supply risk [110] within energy models, these concepts are Many studies are investigating the wind power expansion needs for yet to be widely implemented, and most IAMs and energy models are yet decarbonising the German power system. For example, Fraunhofer ISE to include CRM constraints at all. [124] analysed options for GHG neutrality by 2045 using the REMod The EU seeks to increase its EV fleet from about 3.2 million in 2020 model in which energy system simulation and cost-optimisation are [105] to at least 30 million by 2030 [111]. Although electric car regis­ coupled (hybrid optimisation). Modellers developed four scenarios trations in Europe more than doubled in 2020 compared to 2019 [112], characterised by multiple restrictions. The Reference, Inertia and Suffi­ this target requires a sharp increase in sales, especially as there are large ciency scenarios assume a German onshore wind fleet of up to 230 GW. differences across Europe: Norwegian new sales now exceed 75 %, Meanwhile, in the Unacceptance scenario, where it is assumed that whereas many Eastern and Southern European countries remain below expansion struggles from strong public opposition, capacity only reaches 5 % [113]. Adding to concerns about range and charger availability, up to 80 GW; instead, emissions targets being achieved via a massive many consumers and policymakers also question whether the technol­ expansion of solar PV (660 GW), which may also face strong opposition ogy is more environmentally friendly [114,115]. due to high installation rates. Either way, the more ambitious wind For EV battery production, the materials of most concern in current deployment figures roughly correspond to a tripling and quadrupling of designs are lithium, cobalt and natural graphite. New, advanced battery current onshore capacity and an average expansion pace of up to 7 GW/ designs are under development and may bring increasing demand of year. silicon, titanium and niobium [38], all of which are considered critical The deployment rates required for such climate-neutral system pro­ by the EC [116]. The EU relies almost exclusively on imported raw jections depart strongly from the observed development. While Ger­ materials in battery manufacturing, with one-third imported from China many seemed almost “on track” with growing annual expansion rates and one-fifth from Latin America and Africa, respectively (ibid). Pro­ between 3.7 and 5.2 GW onshore wind power during 2014–2017, this cessed materials, particularly those used for cathodes and anodes, are pace has since dropped to 1.1–1.9 GW/year [125]. The decline was also imported, especially from China (52 %) and Japan (31 %) [116]. caused by changes in the policy support (shift from feed-in-tariff to The EU does not produce any of the finished battery assemblies it uses, auctions) and, especially, difficulties with installation permits, often importing these mainly (66 %) from China (ibid). Europe thus faces a originating from local opposition to new wind power projects 7 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Table 4 Current total EU consumption and projected requirements for EV batteries alone for three key materials. Projected electric vehicle numbers are listed in accordance with low-, medium-, high-demand scenarios (LDS, MDS, HDS). Adapted from Bobba et al. [38]. 2020 2030 2050 LDS MDS HDS LDS MDS HDS Total EU consumption Cobalt 30,000 [tonnes] Lithium 6000 Graphite 250,000 Projected EV requirements Cobalt 38,000 67,000 120,000 38,000 110,000 290,000 [tonnes] Lithium 32,000 51,000 90,000 48,000 130,000 260,000 Graphite 340,000 500,000 820,000 700,000 1,800,000 2,700,000 Projected EV requirements [times total EU consumption in 2020] Cobalt 1.3 2.2 4.0 1.3 3.7 9.7 Lithium 5.3 8.5 15.0 8.0 21.7 43.3 Graphite 1.4 2.0 3.3 2.8 7.2 10.8 [126,127]. Despite broad support for the energy transition in general, change as the government plans to update the expansion targets of one-fifth of the population rejects or strictly rejects further deployment onshore wind power in the Renewable Energy Act, with a substantial of onshore wind power [128], with numerous anti-wind citizen initia­ increase in the annual auction volume to 10 GW by 2027 [134]. If the tives emerging [129]. The causes of opposition, indicated by lawsuits law passes the parliament, Germany is on the path to reach more than and local resistance, are manifold and largely connected to environ­ twice the installed capacity compared to current plans by 2035 and soon mental and social factors. Some 20 % of all onshore wind power projects even overtake the ambitious Reference scenario of the Fraunhofer study. are affected by litigation [63], mainly from environmental organisa­ However, if the government does not react to the causes of the “wind tions, but also from citizens and citizen initiatives, often raising concerns market implosion”, especially the growing opposition to wind power about biodiversity [47]. This situation is similar around Europe plants routed in environmental and social factors, its expansion plans [130,131]. A further key driver of wind power opposition is its land use, might fail and possibly contribute to further resistance. owing to the vast amounts of land required for wind farms, potentially triggering direct land-use conflicts, and public opposition due to visual 4.2.4. Domestic investment behaviour for small-scale PV in Greece and aesthetic landscape impacts [125,126]. In the long run, wind power In June 2009, the Greek government introduced the “Special pro­ land use could be substantial, at least 1–2 % the German land area gramme for the deployment of solar photovoltaics (PV) on buildings and [132,133]. roofs”, which simplified installation procedures for domestic solar PV In summary, large disconnects exist between what models say is installers and provided a generous feed-in-tariff of €550/MWh to attract necessary for carbon-neutrality and what is feasible given the opposition investments [136]. Later that year, the process for the transposition of (Fig. 6). On the one hand, current onshore wind development does not the Renewable Energy Directive into the national legislation was initi­ align with prominent scenario results (except, for example, Unacceptance ated. Different energy models were used in this process to evaluate the scenarios). On the other hand, current policy targets are not ambitious energy policy scenarios developed, and to perform a sensitivity analysis enough to reach the demanded wind fleet. The latter might, however, taking into account different evolution paths of fiscal/regulatory Fig. 6. Installed onshore wind capacity in Germany. Real-world developments, policy goals, and modelled needs in a 100 % renewable future. Data sources: [122,124,134,135] and linear inter- and extrapolation of missing data; rounded values. 8 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 parameters. In particular, the TIMES-MARKAL model [137] was used to optimisation model results ignoring actual behaviour, there is a risk calculate the specific targets for each type of technology, underneath the that the policy will fail, triggering either too much or too little overall national renewable target, while the ENPEP model [138] was investment. used for the assessment of different policy measures for achieving the targets. Both models used inputs from the models WASP [139] (used for 5. Discussion optimum electricity generation planning) and COST (used for the sto­ chastic simulation of the electricity generation system). Based on this Our results show that model users request better integration of the modelling work, the government set the 2020 target to 2200 MW of total environmental and social factors of the energy transition into energy PV capacity [140], while model results suggested that the feed-in-tariff models so that models can provide results that better represent real- policy design would drive consumer investments in a linear way to the world developments, thereby improving their usefulness as policy achievement of the 2020 PV target (Fig. 7). advice tools. According to users, environmental factors should go However, the targets defined by the model and set by the govern­ beyond GHG emissions and include the demand of raw materials, im­ ment were disconnected from the adoption realities on the ground. pacts on biodiversity, land use and water consumption, and other indi­ Many consumers saw the feed-in tariff as an attractive source of addi­ rect and externalised impacts. Among the social factors, we identified tional revenue during a period of great financial distress for the country, social acceptance, individual and community participation, consumer leading to a PV boom in 2009–2013. Thus, the model-based target of behaviour and lifestyles, and policy preferences and dynamics as the 2200 MW of PV capacity by 2020 was met and exceeded in 2013 [141]. most relevant. The identified needs largely align with the environmental Consequently, the government, without consulting any further model- and social factors that are currently discussed in the scientific literature based analyses, imposed an additional tax on consumer incomes from (Section 2), underlining their relevance. renewable electricity generation, simultaneously with a reduction on In our four case studies, we show that omitting the environmental tariffs to counterbalance negative fiscal implications [142]. This politi­ and social factors ranked most important to model users may lead to less cal decision shook the confidence of domestic investors in the stability relevant, or even misleading, results in several ways: and credibility of the support system [143,144], leading to a complete shutdown of the domestic PV market [32]. 1) Neglecting social factors can lead to unrealistic model assumptions Here, once again, neither the political reality nor the consumer and misleading findings about the speed of the transition: The cases of behaviour was reflected in the energy models used to inform policy­ grid expansion in the EU and onshore wind power in Germany show making. Accordingly, model-supported policy expectations and reality that public opposition, related to land-use and biodiversity concerns, diverged: the adoption was first much higher than the energy models substantially delays the implementation of the energy transition, anticipated, and then completely collapsed following the policy change. often by many years and for single projects over a decade. Similarly, Indeed, the policy change was such a strong shock that subsequent ef­ the Greek case study on solar PV illustrates the perils of ignoring forts to rekindle the residential uptake of PV expansion through a net- investor behaviour and solely basing target-setting on top-down metering scheme [146] did not work, causing the updated 2020 PV optimisation models assuming a benevolent central planner, in this target of 3300 MW (by end of 2019) to fail. case resulting in overly rapid deployment. This case study demonstrates the problems arising from a non- 2) Models may make unrealistic assumptions about the potential of adaptive model-informed policymaking process, the consequences of renewable energy if they focus only on the technical potential, not being flexible and allowing for contingency measures in case of a ignoring societal preferences and their impact on land availability. policy failure, and of the necessity for energy models to evaluate con­ The example of onshore wind power in Germany shows that wind sumer response to specific policy incentives, especially when high tariffs power expansion can be hindered when wind turbines are not are provided. By only using top-down optimisation models for target accepted in certain areas. Whether temporary or permanent, such setting, model-informed policymaking risks being misleading: such delays can reduce local and, consequently, countrywide wind ca­ models assume a benevolent planner with central control over the sys­ pacity potentials. tem and investments, but investor behaviour may be very different than 3) Not considering the demand of land and raw materials for renewable this centrally-planned perspective assumes. If policy measures whose energy assets and related infrastructure may generate scenarios that success depends on investor responses are based only on top-down neglect central impacts of the energy transition and support Fig. 7. Total PV capacity installed in Greece during the period 2007–2020: Modelling results vs historical observations. Data source: [140,145]. 9 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 technology options that cannot materialise or that bring substantial improvements [23]. In any case, none of these approaches allow for the supply risks. The case of EV batteries shows that the availability of further analysis of LCA outputs beyond the simple aggregation of values raw materials could become both a deployment constraint and a across system components. geopolitical or economic risk factor as the transition progresses, Few attempts have been made to include detailed information about unless a significant system change is made regarding the recycling of raw materials demand and supply within energy models. One notable materials. Furthermore, land-use conflicts are an increasing problem example is the MEDEAS-World model [152], which includes a module for deployment of renewables and infrastructure. that accounts for the materials and energy required for energy infra­ 4) Ignoring environmental concerns and societal preferences can lead structure manufacturing. The model quantifies the material re­ to problematic or misaligned design of future energy and mobility quirements for implementing renewable energy infrastructure, systems. Our cases indirectly show that consumers have strong atti­ including 19 CRMs, and compares these with current global availability tudes and opinions towards technologies such as EV and wind tur­ estimates to detect potential supply issues. As such, it represents a much- bines and should therefore also have an influence on the siting of needed initial foray into the inclusion of CRM aspects within a detailed renewable energy and planning of solutions more broadly. IAM suite. To facilitate greater analysis of raw material aspects and LCA outputs Collectively, this demonstrates a large need to integrate social and with and across energy system levels, scholars developed the ENBIOS environmental factors as variables in energy models. Achieving this is module [153]. ENBIOS takes system specification data (“energy mix” difficult because such variables must be based on context-specific and other information) from models, combines this data with raw ma­ empirical observation and because transitions are dynamic processes terial requirement information and calculated environmental and socio- and change, possibly fundamentally, over time. For example, environ­ metabolic indicator data to produce extensive outputs such as life cycle mental impacts and resource demands could be reduced during the impact assessment indicators and bespoke indicators derived from life transition via circularity initiatives or by innovations in new materials. cycle and other data. ENBIOS also directly integrates raw material Accordingly, technological change must also be depicted in environ­ supply risks, circularity and local impacts at the point of extraction via a mental assessments, including emerging approaches such as prospective methodology that combines life-cycle inventory data, supply risk and life cycle assessment (LCA). Similarly, the drivers of public opposition end-of-life recycling input rate data [154], and localised environmental against wind power may be different in, for example, Germany and performance data for the countries from which materials are sourced Greece, and will likely change between now and 2030. To depict such [153,154]. As such, it brings a more systemic method to the assessment developments, models may need to endogenise social factors, modelling of material use and environmental impacts than previous approaches the underlying drivers such as regional density and size of existing wind while offering a first attempt at quantifying these impacts alongside the farms [147] instead of considering them as exogenous variables, as in­ socio-metabolic aspects that also apply to energy systems. Outputs from dependent factors of a multi-criteria analysis, or ex-post indicators. ENBIOS can be used to inform the selection of subsequent model sce­ Furthermore, political realities and regulation – both of which are dy­ narios or, for example, guide constraint parameters. namic – greatly affect transitions, for example, by adding constraints such as wind farm distance rules or alleviating constraints such as 5.2. Approaches for integrating social factors in energy models recycling requirements to avoid material shortage. These may also “change the sign” of a factor, turning a barrier into a driver; e.g., The need to integrate social factors into energy models has been enabling community renewables can reduce opposition and create a new previously addressed, and several approaches exist. These are typically potential transition driver. Several approaches for including environ­ focused on implementing social factors as constraints, but some have mental and social factors into energy models are emerging and, although attempted to integrate them as explicit variables within energy models. none of them fully integrate all such issues, the seeds for doing so are For example, the Quantification of Technological DIffusion and sociAl probably being sown; we discuss existing approaches for such integra­ constraiNts (QTDIAN) toolbox allows modellers both to include real- tion in Sections 5.1 and 5.2. world, non-idealised policy constraints (e.g. actual national/regional Adding these additional factors will align energy models more setback distances for wind power), and to base scenario construction on closely to observable realities and will thus make them more policy observed policy objectives beyond GHG elimination, such as decen­ relevant. However, it will also make models more complex, reducing tralisation/centralisation or transmission system preferences [157]. their transparency and risk increasing the “black-box“ nature of models. Seeking to enable model-based assessment of the impact of different To avoid overloading models and, indeed, increase their usefulness as policy measures, Best et al. [158] built a database for energy sufficiency advice tools, it is essential to include stakeholders in the modelling policies, allowing the explicit integration of sufficiency indicators into process, both to provide data and, critically, context for that data and for energy modelling. Presently, there is a strong trend towards integration the context-sensitive interpretation of model outputs. of social science and humanities in energy system analysis, and several model frameworks are being rebuilt to become more realistic and 5.1. Approaches for integrating environmental factors in energy models holistic. Including public acceptance of renewable energy deployment stra­ Several approaches exist to better represent environmental factors in tegies in new modelling frameworks has become a particular recent energy models. One promising approach is the integration of life-cycle focus. One approach is to seek the fair geographical distribution of perspectives and data sources into energy models, which provides production and infrastructure assets, thus avoiding overly strong con­ greater access to high-resolution raw material information and other centration of deployment in single regions [159]. Others seek to valuable environmental indicators. For example, Pehl et al. [148] and generate scenario-based options to identify which parts of a deployment Luderer et al. [149] linked IAMs with LCA information using the THE­ trajectory are necessary, and where more flexibility is available, as a first MIS model [150], enabling high-resolution GHG emissions and several step towards increasing stakeholder engagement and including public other environmental impacts to be included within modelling processes. deliberations about the most attractive pathways for a country or region. This concept has recently been expanded by allowing life-cycle data to For example, the “spatially explicit practically optimal results” be manipulated within Python environments, enabling outputs from (SPORES) approach explores nearly cost-optimal systems. Applying models to be directly automated with LCA calculations. For example, the SPORES to Italy, Lombardi et al. [160] find that only photovoltaic and PREMISE model [151] allows different background electricity mixes and storage technologies are necessary components for a zero‑carbon power other parameters from the IMAGE model to inform future LCA processes system by 2050, whereas wind power choices are more flexible, allow­ to account for future changes in renewable energy use or technological ing for deliberation-centred planning. Yet others include “resistance 10 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 factors” for grid expansion, including these in their model to generate along the overall supply chains that produce energy and infrastructure. delay-minimal expansion pathways instead of purely cost-optimal ones As such, it may be unclear whether the emissions, resource requirements [159]. and impacts assigned to a process are occurring locally or in various Adopting an entirely different approach, McKenna et al. [59] quan­ other regions of the world. This is less of an issue for GHG emissions, tified the visual impacts of onshore wind in energy system analyses, where impacts are assumed to occur globally, regardless of their origin. basing the analysis on “scenicness” values of onshore wind sites [161]. However, a shortcoming exists when assessing more localised In four scenarios for onshore wind potential, they gradually reduced the impacts–e.g., air and water pollution or land use. There are a few ini­ technical potential by quartiles of the scenicness distribution, revealing tiatives working on the regionalization of Life Cycle Data and methods that the windiest locations are generally also the most scenic ones. [166]. This movement is merely taking off, but we do advocate for its Hence, including this parameter in models could greatly reduce the wind relevance and potential contributions in the assessment of environ­ power potential, while generating more relevant results and exposing mental (and social) impacts of the energy transition. conflicts between landscape protection and renewables, facilitating Lastly, we call for more research on integrating or linking environ­ solution-oriented deliberation. mental and social aspects into and with modelling, going beyond the Finally, although energy modelling is still dominated by central consideration of those factors as “add-ons”. For example, future research planner-based optimisation modelling, alternatives are emerging, could soft link energy models with environmental models to assess wider including models that describe actor behaviour instead of top-down environmental impacts of transition pathways and energy systems. optimal deployment [88,162]. Such models, including ABMs, can be used both to inform policy design decisions and to set appropriate tar­ 6. Conclusions gets. For example, Melliger et al. [163] explored the effects of exposing renewable electricity technologies to market competition using an ABM We conclude that users want better models regarding the represen­ fed with investor behaviour data from a conjoint analysis [164]. They tation of environmental and social factors, and we have demonstrated show that although policies to increase competition seek to reduce en­ that ignoring these critical aspects of the energy transition can lead to ergy system costs, they likely both slow down deployment and increase wrong or misleading evidence about the potential of renewable elec­ costs because investors flock to still supported and more expensive tricity, and the speed, impacts and technological options of the energy technologies. Similarly, addressing the same case in section 4.2.4, the transition. While the modelling community is taking steps to better agent-based technology adoption model (ATOM) simulates the diffusion incorporate social and environmental factors into energy models, our of small-scale PV in Greece under the net-metering scheme currently in results suggest that many of these key areas are not yet considered in operation [30,141], based on behavioural profiles of small-scale in­ sufficient detail and that existing approaches have not been sufficiently vestors. Indeed, ATOM shows that the existing net-metering policy is applied. While energy models will undoubtedly continue to be used to unable to achieve the 2025 and 2030 PV targets due to policy short­ inform policymaking, our study is a call to energy modellers to further comings, a finding that could not be detected using system optimisation advance the representation of these factors in models or to advance the models. interlinking of different modelling tools. This includes the main­ streaming of social and environmental factors as explicit variables in 5.3. Limitations and future research models, possibly even by endogenising particularly important parame­ ters, such as social preferences, into the models based on context- Our study shows that energy models should consider environmental sensitive empirical data. Including these factors would vastly improve and social aspects of the energy transition and indicates the magnitude the robustness of energy system models and, ultimately, would increase of the problems arising by ignoring such factors. However, we the suitability and meaningfulness of models to informing policy deci­ acknowledge the limitations of our study in terms of generalisability of sion regarding the complex interplay between energy requirements, user needs as we were not able to capture needs and differences between societal objectives and environmental considerations as Europe and the EU member states, or in countries beyond the EU. Just as social and world continue advancing towards climate neutrality. environmental barriers and drivers may differ across both time and countries, user needs for model-based information will also be both dynamic and context-sensitive. Whereas we believe that the barriers we Declaration of competing interest explore – material requirements, opposition, etc. – are relevant to all countries, the relative importance of each factor may differ, depending The authors declare that they have no known competing financial on political factors, geography and transition progress. Further studies interests or personal relationships that could have appeared to influence could explore the context-specific needs and reasons for modelling re­ the work reported in this paper. quirements to support the further improvement of modelling tools tailored to specific countries and challenges. Data availability Furthermore, we have presented illustrative, non-exclusive examples of situations where energy modelling studies have generated problem­ Data will be made available on request. atic or unfeasible policy-advice because they did not sufficiently consider environmental or social factors. However, model results do not Acknowledgements (always) directly lead to policy decisions, but are – and should be – only one source of information among others [165]. Further research could This work has received funding from the European Union's Horizon investigate how policy and decision-making processes deal with factors 2020 research and innovation programme (Grant agreement 837089, that are not considered in energy models and what concrete impact this SENTINEL project). We would also like to thank the stakeholders for has on policy decisions. taking the time to participate in the online interviews, the online survey, It is also noted that the levels of granularity within current LCA and the workshop. This paper is an original paper, building on content databases make it difficult to localise the various impacts that occur from the SENTINEL Deliverables 1.2, 2.1 and 2.4 by the same authors. 11 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Appendix A Table A1 . Stakeholder engagement activities and participants. Method Questions and content Engaged model users Further information Interviews in five jurisdictions: The interview guidelines included these questions to energy 32 interviewees: 11 policymakers, 4 energy industry Complete interview the EU, Germany, Greece, modellers: experts, 5 NGO representatives, 12 researchers guideline: [167] Poland, Sweden - In your opinion, what kind of information should an energy model deliver, now and in the future, to inform decision- making (processes) in energy policy? - In your opinion, how should the process of model development be designed to increase the chance of the later model use in policymaking? - Which conditions must be given that increase the chance that you would use the models or the results, respectively, in future policymaking/your work? Question to non-modellers included: - What are the current and future challenges or aspects of the energy transition that should be integrated into future energy models? - In your opinion, what kind of information should an energy model deliver to help make good decisions about energy policy/energy issues? - Which conditions must be given that increase the chance that you would use the models or the results, respectively, in future policymaking/your work? Online survey The survey was designed around six sections: 1) Personal 90 complete responses from 12 policymakers, 16 Complete survey: background, model use, and general demands, 2) Model energy industry experts, 11 NGO representatives, 42 [91] content, 3) Model design, 4) Modelling process, 5) Modelling researchers outreach, 6) Others and demographic data. Workshop on user needs for Five breakout sessions: 23 non-SENTINEL participants from different European Full list of energy modelling, European Session 1: Social and policy aspects in energy models Member States: 2 policymakers, 11 energy industry participants: [168] Member States Session 2: Including environmental aspects in energy system experts, 6 NGO representatives, 4 researchers models Session 3: Modelling energy demand and supply Session 4: Modelling the economic impacts of the energy transition Session 5: Designing the model platform of SENTINEL 12 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Fig. A1. Graphical recording of the workshop, including a ranking of factors that receive more attention in energy models. References [11] B.K. Sovacool, D.J. Hess, R. Cantoni, D. Lee, M. Claire Brisbois, H. Jakob Walnum, R. Freng Dale, B. Johnsen Rygg, M. Korsnes, A. Goswami, S. Kedia, S. Goel, Conflicted transitions: exploring the actors, tactics, and outcomes of social [1] European Commission, Communication from the Commission to the European opposition against energy infrastructure, Glob. Environ. Chang. 73 (2022), Parliament, the European Council, the Council, the European Economic and 102473, https://doi.org/10.1016/j.gloenvcha.2022.102473. Social Committee and the Committee of the Regions: The European Green Deal - [12] G. Savvidis, K. Siala, C. Weissbart, L. Schmidt, F. Borggrefe, S. Kumar, K. Pittel, COM/2019/640 Final, European Commission, Brussels, 2019. R. Madlener, K. Hufendiek, The gap between energy policy challenges and model [2] N. Martin, C. Madrid-López, L. Talens-Peiró, D. Süsser, H. Gaschnig, J. Lilliestam, capabilities, Energy Policy 125 (2019) 503–520, https://doi.org/10.1016/j. Observed trends and modelling paradigms on the social and environmental enpol.2018.10.033. aspects of the energy transition, in: Deliverable 2.1. Sustainable Energy [13] S. Pfenninger, A. Hawkes, J. Keirstead, Energy systems modeling for twenty-first Transitions Laboratory (SENTINEL) Project (1.0), 2020, https://doi.org/ century energy challenges, Renew. Sust. Energ. Rev. 33 (2014) 74–86, https:// 10.5281/zenodo.4917183. doi.org/10.1016/j.rser.2014.02.003. [3] European Commission, Communication from the Commission to the European [14] S. Chatterjee, V. Stavrakas, G. Oreggioni, D. Süsser, I. Staffell, J. Lilliestam, Parliament, the Council, the European Economic and Social Committee, the G. Molnar, A. Flamos, D. Ürge-Vorsatz, Existing tools, user needs and required Committee of the Regions and the European Investment Bank, A Framework model adjustments for energy demand modelling of a carbon-neutral Europe, Strategy for a Resilient Energy Union with a Forward-Looking, 2015. Energy Res. Soc. Sci. 90 (2022), 102662, https://doi.org/10.1016/j. [4] European Commission, in: Clean Energy for all Europeans COM(2016) 860 Final, erss.2022.102662. 2016, pp. 1–13. https://eur-lex.europa.eu/resource.html?uri=cellar:fa6ea15b-b7 [15] G. Savvidis, K. Siala, C. Weissbart, L. Schmidt, F. Borggrefe, S. Kumar, K. Pittel, b0-11e6-9e3c-01aa75ed71a1.0001.02/DOC_1&format=PDF. R. Madlener, K. Hufendiek, The gap between energy policy challenges and model [5] IAMC, Integrated Assessment Modeling Consortium Wiki, 2020. capabilities, Energy Policy 125 (2019) 503–520, https://doi.org/10.1016/j. [6] A. Krumm D. Süsser P. Blechinger , Modelling social aspects of the energy enpol.2018.10.033. transition: current and potential representations in energy models, Energy. (n.d.). [16] R.H.E.M. Koppelaar, J. Keirstead, N. Shah, J. Woods, A review of policy analysis [7] E. Trutnevyte, L.F. Hirt, N. Bauer, A. Cherp, A. Hawkes, O.Y. Edelenbosch, purpose and capabilities of electricity system models, Renew. Sust. Energ. Rev. 59 S. Pedde, D.P. van Vuuren, Societal transformations in models for energy and (2016) 1531–1544, https://doi.org/10.1016/j.rser.2016.01.090. climate policy: the ambitious next step, One Earth 1 (2019) 423–433, https://doi. [17] C.A. Scott, S.A. Pierce, M.J. Pasqualetti, A.L. Jones, B.E. Montz, J.H. Hoover, org/10.1016/j.oneear.2019.12.002. Policy and institutional dimensions of the water-energy nexus, Energy Policy 39 [8] A. Nikas, J. Lieu, A. Sorman, A. Gambhir, E. Turhan, B.V. Baptista, H. Doukas, (2011) 6622–6630, https://doi.org/10.1016/j.enpol.2011.08.013. The desirability of transitions in demand: incorporating behavioural and societal [18] M.M. Sokołowski, R.J. Heffron, Defining and conceptualising energy policy transformations into energy modelling, Energy Res. Soc. Sci. 70 (2020), 101780, failure: the when, where, why, and how, Energy Policy 161 (2022), 112745, https://doi.org/10.1016/j.erss.2020.101780. https://doi.org/10.1016/J.ENPOL.2021.112745. [9] V. Moreau, P.C.Dos Reis, F. Vuille, Enough metals? Resource constraints to supply [19] D. Süsser, H. Gaschnig, A. Ceglarz, V. Stavrakas, A. Flamos, J. Lilliestam, Better a fully renewable energy system, Resources 8 (2019) 29, https://doi.org/ suited or just more complex? On the fit between user needs and modeller-driven 10.3390/resources8010029. improvements of energy system models, Energy (2021), 121909, https://doi.org/ [10] G. Calvo, A. Valero, Strategic mineral resources: availability and future 10.1016/j.energy.2021.121909. estimations for the renewable energy sector, Environ. Dev. (2021), 100640, https://doi.org/10.1016/j.envdev.2021.100640. 13 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 [20] J.A. Hollingsworth, E. Ravishankar, B. O’Connor, J.X. Johnson, J.F. DeCarolis, [45] A. Mayer, W. Haas, D. Wiedenhofer, F. Krausmann, P. Nuss, G.A. Blengini, Environmental and economic impacts of solar-powered integrated greenhouses, Measuring progress towards a circular economy: a monitoring framework for J. Ind. Ecol. 24 (2020) 234–247, https://doi.org/10.1111/jiec.12934. economy-wide material loop closing in the EU28, J. Ind. Ecol. 23 (2019) 62–76, [21] J. Li, S. Li, F. Wu, Research on carbon emission reduction benefit of wind power https://doi.org/10.1111/jiec.12809. project based on life cycle assessment theory, Renew. Energy 155 (2020) [46] F. Diaz-Maurin, J.J. Cadillo-Benalcazar, Z. Kovacic, C. Madrid-López, T. Serrano- 456–468, https://doi.org/10.1016/j.renene.2020.03.133. Tovar, M. Giampietro, R.J. Aspinall, J. Ramos-Martin, S.G.F. Bukkens, The [22] C.J. Barnhart, S.M. Benson, in: On the Importance of Reducing the Energetic and Republic of South Africa, in: M. Giampietro, R.J. Aspinall, J. Ramos-Martin, S.G. Material Demands of Electrical Energy Storage 6, 2013, pp. 1083–1092, https:// F. Bukkens (Eds.), Resour. Account. Sustain. Nexus Between Energy, Food, Water doi.org/10.1039/C3EE24040A. L. Use, Routledge, Abingdon, 2014, pp. 194–213, https://doi.org/10.4324/ [23] A. Mendoza Beltran, B. Cox, C. Mutel, D.P. van Vuuren, D. Font Vivanco, 9781315866895. S. Deetman, O.Y. Edelenbosch, J. Guinée, A. Tukker, D.P. Vuuren, D. Font [47] C.C. Voigt, T.M. Straka, M. Fritze, Producing wind energy at the cost of Vivanco, S. Deetman, O.Y. Edelenbosch, J. Guinée, A. Tukker, When the biodiversity: a stakeholder view on a green-green dilemma, J. Renew. Sustain. background matters: using scenarios from integrated assessment models in Energy 11 (2019), 063303, https://doi.org/10.1063/1.5118784. prospective life cycle assessment, J. Ind. Ecol. 24 (2020) 64–79, https://doi.org/ [48] L. Bennun, J. van Bochove, C. Ng, C. Fletcher, D. Wilson, N. Phair, G. Carbone, 10.1111/jiec.12825. Mitigating biodiversity impacts associated with solar and wind energy [24] T. Watari, B.C. McLellan, D. Giurco, E. Dominish, E. Yamasue, K. Nansai, Total development, in: Guidelines for Project Developers, IUCN, Gland, 2021, https:// material requirement for the global energy transition to 2050: a focus on doi.org/10.2305/IUCN.CH.2021.04.en. transport and electricity, Resour. Conserv. Recycl. 148 (2019) 91–103, https:// [49] A. Santangeli, T. Toivonen, F.M. Pouzols, M. Pogson, A. Hastings, P. Smith, doi.org/10.1016/j.resconrec.2019.05.015. A. Moilanen, Global change synergies and trade-offs between renewable energy [25] S. Nabernegg, B. Bednar-Friedl, P. Muñoz, M. Titz, J. Vogel, National policies for and biodiversity, GCB Bioenergy 8 (2016) 941–951, https://doi.org/10.1111/ global emission reductions: effectiveness of carbon emission reductions in gcbb.12299. international supply chains, Ecol. Econ. 158 (2019) 146–157, https://doi.org/ [50] A. Gasparatos, C.N.H. Doll, M. Esteban, A. Ahmed, T.A. Olang, Renewable energy 10.1016/j.ecolecon.2018.12.006. and biodiversity: implications for transitioning to a green economy, Renew. Sust. [26] F. Lombardi, M.V. Rocco, E. Colombo, A multi-layer energy modelling Energ. Rev. 70 (2017) 161–184, https://doi.org/10.1016/j.rser.2016.08.030. methodology to assess the impact of heat-electricity integration strategies: the [51] M. Giamalaki, T. Tsoutsos, Sustainable siting of solar power installations in case of the residential cooking sector in Italy, Energy 170 (2019) 1249–1260, Mediterranean using a GIS/AHP approach, Renew. Energy 141 (2019) 64–75, https://doi.org/10.1016/j.energy.2019.01.004. https://doi.org/10.1016/j.renene.2019.03.100. [27] T. Boßmann, I. Staffell, The shape of future electricity demand: exploring load [52] S. Gross, Renewables, Land Use, and Local Opposition in the United States, curves in 2050s Germany and Britain, Energy 90 (2015) 1317–1333, https://doi. Brookings, Washington, DC, 2020. org/10.1016/j.energy.2015.06.082. [53] G. Felber, G. Stoeglehner, Onshore wind energy use in spatial planning—a [28] V. Stavrakas, A. Flamos, A modular high-resolution demand-side management proposal for resolving conflicts with a dynamic safety distance approach, Energy. model to quantify benefits of demand-flexibility in the residential sector, Energy Sustain. Soc. 4 (2014) 22, https://doi.org/10.1186/s13705-014-0022-8. Convers. Manag. 205 (2020), 112339, https://doi.org/10.1016/j. [54] K. Huang, M.J. Eckelman, Appending material flows to the National Energy enconman.2019.112339. Modeling System (NEMS) for projecting the physical economy of the United [29] T. Tröndle, J. Lilliestam, S. Marelli, S. Pfenninger, Trade-offs between geographic States, J. Ind. Ecol. (2020), https://doi.org/10.1111/jiec.13053. scale, cost, and infrastructure requirements for fully renewable electricity in [55] E. Rinne, H. Holttinen, J. Kiviluoma, S. Rissanen, Effects of turbine technology Europe, Joule 4 (2020) 1929–1948, https://doi.org/10.1016/j. and land use on wind power resource potential, Nat. Energy 3 (2018) 494–500, joule.2020.07.018. https://doi.org/10.1038/s41560-018-0137-9. [30] V. Stavrakas, S. Papadelis, A. Flamos, An agent-based model to simulate [56] R.Y. Shum, A comparison of land-use requirements in solar-based technology adoption quantifying behavioural uncertainty of consumers, Appl. decarbonization scenarios, Energy Policy 109 (2017) 460–462, https://doi.org/ Energy 255 (2019), 113795, https://doi.org/10.1016/j.apenergy.2019.113795. 10.1016/j.enpol.2017.07.014. [31] T. Perger, L. Wachter, A. Fleischhacker, H. Auer, PV sharing in local communities: [57] I. Capellán-Pérez, C. de Castro, I. Arto, Assessing vulnerabilities and limits in the peer-to-peer trading under consideration of the prosumers’ willingness-to-pay, transition to renewable energies: land requirements under 100% solar energy Sustain. Cities Soc. (2020), 102634, https://doi.org/10.1016/j.scs.2020.102634. scenarios, Renew. Sust. Energ. Rev. 77 (2017) 760–782, https://doi.org/ [32] A. Nikas, V. Stavrakas, A. Arsenopoulos, H. Doukas, M. Antosiewicz, 10.1016/j.rser.2017.03.137. J. Witajewski-Baltvilks, A. Flamos, Barriers to and consequences of a solar-based [58] K. Palmer-Wilson, J. Donald, B. Robertson, B. Lyseng, V. Keller, M. Fowler, energy transition in Greece, Environ. Innov. Soc. Trans. 35 (2019) 383–399, C. Wade, S. Scholtysik, P. Wild, A. Rowe, Impact of land requirements on https://doi.org/10.1016/j.eist.2018.12.004. electricity system decarbonisation pathways, Energy Policy 129 (2019) 193–205, [33] D.J. Hess, R.G. McKane, C. Pietzryk, End of the Line: Environmental Justice, https://doi.org/10.1016/j.enpol.2019.01.071. Energy Justice, and Opposition to Power Lines, 2021, https://doi.org/10.1080/ [59] R. McKenna, I. Mulalic, I. Soutar, J.M. Weinand, J. Price, S. Petrović, K. Mainzer, 09644016.2021.1952799 doi:10.1080/09644016.2021.1952799. Exploring trade-offs between landscape impact, land use and resource quality for [34] S. Avila, Environmental justice and the expanding geography of wind power onshore variable renewable energy: an application to Great Britain, Energy conflicts, Sustain. Sci. 133 (13) (2018) 599–616, https://doi.org/10.1007/ (2022), 123754, https://doi.org/10.1016/j.energy.2022.123754. S11625-018-0547-4, 2018. [60] European Union, Climate Change. Special Eurobarometer 513, 2021, https://doi. [35] G. Iacobuta, N.K. Dubash, P. Upadhyaya, M. Deribe, N. Höhne, National climate org/10.2834/437. change mitigation legislation, strategy and targets: a global update, Clim. Policy [61] J. Cohen, K. Moeltner, J. Reichl, M. Schmidthaler, An empirical analysis of local 18 (2018) 1114–1132, https://doi.org/10.1080/14693062.2018.1489772. opposition to new transmission lines across the EU-27, Energy J. 37 (2016), [36] M. Pehl, A. Arvesen, F. Humpenöder, A. Popp, E.G. Hertwich, G. Luderer, https://doi.org/10.5547/01956574.37.3.jcoh. Understanding future emissions from low-carbon power systems by integration of [62] P. Devine-Wright, S. Batel, My neighbourhood, my country or my planet? The life-cycle assessment and integrated energy modelling, Nat. Energy 2 (2017) influence of multiple place attachments and climate change concern on social 939–945, https://doi.org/10.1038/s41560-017-0032-9. acceptance of energy infrastructure, Glob. Environ. Chang. 47 (2017) 110–120, [37] S. Carrara, P.Alves Dias, B. Plazzotta, C. Pavel, Raw Materials Demand for Wind https://doi.org/10.1016/j.gloenvcha.2017.08.003. and Solar PV Technologies in the Transition Towards a Decarbonised Energy [63] J. Quentin, Hemmnisse beim Ausbau der Windenergie in Deutschland. Ergebnisse System, EUR 30095 EN, Publications Office of the European Union, Luxembourg, einer Branchenumfrage zu Klagen gegen Windenergieanlagen sowie zu 2020, https://doi.org/10.2760/160859. Genehmigungshemmnissen durch Drehfunkfeuer und militärische Belange der [38] S. Bobba, S. Carrara, J. Huisman, F. Mathieux, C. Pavel, Critical Raw Materials for Luftraumnutzung, Berlin, 2019. Strategic Technologies and Sectors in the EU: A Foresight Study, Publications [64] K. Borch, Mapping value perspectives on wind power projects: the case of the Office of the European Union, Luxembourg, 2020, https://doi.org/10.2873/ danish test Centre for large wind turbines, Energy Policy 123 (2018) 251–258, 865242. https://doi.org/10.1016/j.enpol.2018.08.056. [40] L. Liu, G. Huang, B. Baetz, K. Zhang, Environmentally-extended input-output [65] A. Månsson, A resource curse for renewables? Conflict and cooperation in the simulation for analyzing production-based and consumption-based industrial renewable energy sector, Energy Res. Soc. Sci. 10 (2015) 1–9, https://doi.org/ greenhouse gas mitigation policies, Appl. Energy 232 (2018) 69–78, https://doi. 10.1016/j.erss.2015.06.008. org/10.1016/j.apenergy.2018.09.192. [66] C.C. Voigt, T.M. Straka, M. Fritze, Producing wind energy at the cost of [41] G. Saevarsdottir, H. Kvande, B.J. Welch, Aluminum production in the times of biodiversity: a stakeholder view on a green-green dilemma, J. Renew. Sustain. climate change: the global challenge to reduce the carbon footprint and prevent Energy 11 (2019), 063303, https://doi.org/10.1063/1.5118784. carbon leakage, JOM 72 (2020) 296–308, https://doi.org/10.1007/s11837-019- [67] D.P. Vasilakis, D.P. Whitfield, S. Schindler, K.S. Poirazidis, V. Kati, Reconciling 03918-6. endangered species conservation with wind farm development: cinereous [42] É. Lèbre, M. Stringer, K. Svobodova, J.R. Owen, D. Kemp, C. Côte, A. Arratia- vultures (Aegypius monachus) in South-Eastern Europe, Biol. Conserv. 196 Solar, R.K. Valenta, The social and environmental complexities of extracting (2016) 10–17, https://doi.org/10.1016/j.biocon.2016.01.014. energy transition metals, Nat. Commun. 111 (11) (2020) 1–8, https://doi.org/ [68] V. Kati, C. Kassara, Z. Vrontisi, A. Moustakas, The biodiversity-wind energy-land 10.1038/s41467-020-18661-9, 2020. use nexus in a global biodiversity hotspot, Sci. Total Environ. 768 (2021), [43] J. Ma, Q. Duan, Environmental dumping and international unionized oligopolies, 144471, https://doi.org/10.1016/j.scitotenv.2020.144471. SSRN Electron. J. (2009), https://doi.org/10.2139/ssrn.1494877. [69] L.D. Knopper, C.A. Ollson, L.C. McCallum, M.L.W. Aslund, R.G. Berger, [44] G. Gaustad, M. Krystofik, M. Bustamante, K. Badami, Circular economy strategies K. Souweine, M. McDaniel, Wind turbines and human health, Front. Public Health for mitigating critical material supply issues, Resour. Conserv. Recycl. 135 (2018) 2 (2014) 1–20, https://doi.org/10.3389/fpubh.2014.00063. 24–33, https://doi.org/10.1016/j.resconrec.2017.08.002. 14 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 [70] F.W. Geels, Socio-technical transitions to sustainability: a review of criticisms and [95] Energinet, Kassø-Frøslev: New Electricity Interconnector to Germany, 2020. elaborations of the multi-level perspective, Curr. Opin. Environ. Sustain. (2019), [96] J. Lilliestam, S. Ellenbeck, C. Karakosta, N. Caldés, Understanding the absence of https://doi.org/10.1016/j.cosust.2019.06.009. In press. renewable electricity imports to the European Union, Int. J. Energy Sect. Manag. [71] R. Cowell, G. Bristow, M. Munday, Acceptance, acceptability and environmental 10 (2016) 291–311, https://doi.org/10.1108/IJESM-10-2014-0002. justice: the role of community benefits in wind energy development, J. Environ. [97] European Commission, EU Reference Scenario 2020. Energy, Transport and GHG Plan. Manag. 54 (2011) 539–557, https://doi.org/10.1080/ Emissions. Trends to 2050, 2021. 09640568.2010.521047. [98] ENTSO-E, Ten-Year Network Development Plan 2020, 2021. [72] D. Süsser, A. Kannen, ‘Renewables? Yes, please!’: perceptions and assessment of [99] ENTSO-E, TYNDP 2020 Projects Sheets, 2021. community transition induced by renewable-energy projects in north frisia, [100] ENTSO-E, in: 10-Year Network Development Plan 2012, Report, 2012, Sustain. Sci. 12 (2017) 563–578, https://doi.org/10.1007/s11625-017-0433-5. pp. 1–219. [73] B.J.A. Walker, B. Wiersma, E. Bailey, Community benefits, framing and the social [101] G.K. Pall, A.J. Bridge, J. Gray, M. Skitmore, Causes of delay in power transmission acceptance of offshore wind farms: an experimental study in England, Energy Res. projects: an empirical study, Energies 13 (2019) 17, https://doi.org/10.3390/ Soc. Sci. 3 (2014) 46–54, https://doi.org/10.1016/j.erss.2014.07.003. en13010017. [74] J. Balest, E. Pisani, D. Vettorato, L. Secco, Local reflections on low-carbon energy [102] S. Perras, Electricity Transmission Line Planning - Success Factors for systems: a systematic review of actors, processes, and networks of local societies, Transmission System Operators to Reduce Public Opposition, Techn. Univ., Diss, Energy Res. Soc. Sci. 42 (2018) 170–181, https://doi.org/10.1016/j. Dresden, 2015. erss.2018.03.006. [103] L.D. Biasotto, A. Kindel, Power lines and impacts on biodiversity: a systematic [75] European Commission, in: Energy Union Package - A Framework Strategy for a review, Environ. Impact Assess. Rev. 71 (2018) 110–119, https://doi.org/ Resilient Energy Union with a Forward-Looking Climate Change Policy, COM 10.1016/j.eiar.2018.04.010. (2015) 80 Final, 2015, pp. 1–21. [104] European Commission, The European Green Deal, 2019, https://doi.org/ [76] J. Stephenson, B. Barton, G. Carrington, A. Doering, R. Ford, D. Hopkins, 10.1128/AAC.03728-14. R. Lawson, A. McCarthy, D. Rees, M. Scott, P. Thorsnes, S. Walton, J. Williams, [105] IEA, Global EV Outlook 2021, International Energy Agency, Paris, 2021. B. Wooliscroft, The energy cultures framework: exploring the role of norms, [106] IEA, Global EV data explorer. https://www.iea.org/articles/global-ev-data practices and material culture in shaping energy behaviour in New Zealand, -explorer, 2021. Energy Res. Soc. Sci. 7 (2015) 117–123, https://doi.org/10.1016/j. [107] S. Statharas, Y. Moysoglou, P. Siskos, G. Zazias, P. Capros, Factors influencing erss.2015.03.005. electric vehicle penetration in the E.U. By 2030: a model-based policy assessment, [77] F. Creutzig, J. Roy, W.F. Lamb, I.M.L. Azevedo, W. Bruine De Bruin, H. Dalkmann, Energies 12 (2019) 2739, https://doi.org/10.3390/en12142739. O.Y. Edelenbosch, F.W. Geels, A. Grubler, C. Hepburn, E.G. Hertwich, R. Khosla, [108] C. Xu, Q. Dai, L. Gaines, M. Hu, A. Tukker, B. Steubing, Future material demand L. Mattauch, J.C. Minx, A. Ramakrishnan, N.D. Rao, J.K. Steinberger, M. Tavoni, for automotive lithium-based batteries, Commun. Mater. 1 (2020) 99, https://doi. D. Ürge-Vorsatz, E.U. Weber, Towards demand-side solutions for mitigating org/10.1038/s43246-020-00095-x. climate change, Nat. Clim. Chang. 8 (2018) 268–271, https://doi.org/10.1038/ [109] A. Boubault, S. Kang, N. Maïzi, Closing the TIMES integrated assessment model s41558-018-0121-1. (TIAM-FR) raw materials gap with life cycle inventories, J. Ind. Ecol. 23 (2019) [78] E. Toulouse, H. Gorge, M. Le Dû, L. Semal, Stimulating energy sufficiency: 587–600, https://doi.org/10.1111/jiec.12780. barriers and opportunities, ECEEE Summer Study Proc. (2017) 59–68. [110] L. Talens-Peiró, N. Martin, G. Villalba-Méndez, C. Madrid-López, Integration of [79] S. Samadi, M.-C. Gröne, U. Schneidewind, H.-J. Luhmann, J. Venjakob, B. Best, raw materials indicators of energy technologies into energy system models, Appl. Sufficiency in energy scenario studies: taking the potential benefits of lifestyle Energy 307 (2022), 118150, https://doi.org/10.1016/j.apenergy.2021.118150. changes into account, Technol. Forecast. Soc. Chang. 124 (2017) 126–134, [111] European Committee of the Regions, EU Must Support Every Region, City and https://doi.org/10.1016/j.techfore.2016.09.013. Village to Deliver Zero Transport Emissions by 2050, 2022 (n.d.). [80] T. Göllinger, Systemisches Innovations- und Nachhaltigkeitsmanagement, 2012. [112] IEA, Trends and Developments in Electric Vehicle Markets, 2021. [81] J.E. Morrissey, S. Axon, R. Aiesha, J. Hillman, A. Revez, B. Lennon, N.P. Dunphy, [113] EEA, New registrations of electric vehicles in Europe. https://www.eea.europa. M. Salel, E. Boo, Identification and characterisation of energy behaviour change eu/ims/new-registrations-of-electric-vehicles, 2021. initiatives, in: Deliverable D4.4 of ENTRUST Project, 2016, https://doi.org/ [114] A. Burkert, H. Fechtner, B. Schmuelling, Interdisciplinary analysis of social 10.5281/ZENODO.3479377. acceptance regarding electric vehicles with a focus on charging infrastructure and [82] C. Zell-Ziegler, J. Thema, B. Best, F. Wiese, J. Lage, A. Schmidt, E. Toulouse, driving range in Germany, World Electr. Veh. J. 12 (2021) 25, https://doi.org/ S. Stagl, Enough? The role of sufficiency in European energy and climate plans, 10.3390/wevj12010025. Energy Policy 157 (2021), 112483, https://doi.org/10.1016/j. [115] Continental, Many People Still Doubtful About Electric Cars’ Environmental enpol.2021.112483. Friendliness, 2021. [83] L. Tummers, Public policy and behavior change, Public Adm. Rev. 79 (2019) [116] European Commission, Study on the EU’s List of Critical Raw Materials (2020) - 925–930, https://doi.org/10.1111/puar.13109. Final Report, Publications Office of the European Union, Luxembourg, 2020, [84] L.V. Lerman, W. Gerstlberger, M.Ferreira Lima, A.G. Frank, How governments, https://doi.org/10.2873/11619. universities, and companies contribute to renewable energy development? A [117] M. Naumanen, T. Uusitalo, E. Huttunen-Saarivirta, R. van der Have, Development municipal innovation policy perspective of the triple helix, Energy Res. Soc. Sci. strategies for heavy duty electric battery vehicles: comparison between China, 71 (2021), 101854, https://doi.org/10.1016/j.erss.2020.101854. EU, Japan and USA, Resour. Conserv. Recycl. 151 (2019), 104413, https://doi. [85] F. Mey, M. Diesendorf, Who owns an energy transition? Strategic action fields and org/10.1016/j.resconrec.2019.104413. community wind energy in Denmark, Energy Res. Soc. Sci. 35 (2018) 108–117, [118] L. Xu, M. Fuss, W.R. Poganietz, P. Jochem, S. Schreiber, C. Zoephel, N. Brown, An https://doi.org/10.1016/j.erss.2017.10.044. environmental assessment framework for energy system analysis (EAFESA): the [86] D. Süsser, M. Döring, B.M.W. Ratter, Harvesting energy: place and local method and its application to the european energy system transformation, entrepreneurship in community-based renewable energy transition, Energy Policy J. Clean. Prod. 243 (2020), 118614, https://doi.org/10.1016/j. 101 (2017) 332–341, https://doi.org/10.1016/j.enpol.2016.10.018. jclepro.2019.118614. [87] T. Perger, L. Wachter, A. Fleischhacker, H. Auer, PV sharing in local communities: [119] Statista, Anteil Erneuerbarer Energieträger an der Bruttostromerzeugung in peer-to-peer trading under consideration of the prosumers’ willingness-to-pay, Deutschland in den Jahren 2020 und 2021. https://de.statista.com/statistik/dat Sustain. Cities Soc. (2020), 102634, https://doi.org/10.1016/j.scs.2020.102634. en/studie/171368/umfrage/struktur-der-bruttostromerzeugung-durch-erneuer [88] A. Mittal, C.C. Krejci, M.C. Dorneich, D. Fickes, An agent-based approach to bare-energien-in-deutschland/, 2022. modeling zero energy communities, Sol. Energy 191 (2019) 193–204, https://doi. [120] Arbeitsgruppe Erneuerbare Energien - Statistik, Monatsbericht zur Entwicklung org/10.1016/j.solener.2019.08.040. der erneuerbaren Stromerzeugung und Leistung in Deutschland. Stand: [89] D. Süsser, H. Gaschnig, A. Ceglarz, V. Stavrakas, A. Flamos, J. Lilliestam, Better 12.06.2020, Umweltbundesamt, 2020. suited or just more complex? On the fit between user needs and modeller-driven [121] Fraunhofer ISI, Consentec, ifeu, Langfristszenarien für die Transformation des improvements of energy system models, Energy 239 (2022), 121909, https://doi. Energiesystems in Deutschland. Modul 0: Zentrale Ergebnisse und org/10.1016/j.energy.2021.121909. Schlussfolgerungen Studie im Auftrag des Bundesministeriums für Wirtschaft und [90] D. Süsser, A. Ceglarz, V. Stavrakas, J. Lilliestam, COVID-19 vs. Stakeholder Energie, 2017. engagement: the impact of coronavirus containment measures on stakeholder [122] Bundesregierung, in: Gesetz für den Ausbau erneuerbarer Energien involvement in european energy research projects, Open Res. Eur. 1 (2021) 57, (ErneuerbareEnergien-Gesetz - EEG 2021), 2021, p. 3138. Manuscript. https://doi.org/10.12688/openreseurope.13683.1. [123] Das Umweltbundesamt, Windenergie an Land. https://www.umweltbundesamt. [91] H. Gaschnig, D. Süsser, A. Ceglarz, V. Stavrakas, A. Flamos, J. Lilliestam, Survey de/themen/klima-energie/erneuerbare-energien/windenergie-an-land#flaeche, Questionnaire and Results on User Needs for Energy Models for the European 2021. Energy Transition, Related to Süsser et al. (2021), Zenodo, 2021 [Data set]. [124] J. Brandes, M. Haun, D. Wrede, P. Jürgens, C. Kost, H.-M. Henning, Wege zu [92] H. Gaschnig, D. Süsser, A. Ceglarz, G. Stavrakas, V. Giannakidis, A. Flamos, einem klimaneutralen Energiesystem – Update Klimaneutralität 2045, 2021. A. Sander, J. Lilliestam, User needs for an energy system modeling platform for [125] Deutsche WindGuard GmbH, Status des Windenergiezubaus an Land in the European energy transition. Deliverable 1.2. Sustainable Energy Transitions Deutschland - Halbjahr 2021, 2021. Laboratory (SENTINEL) project, in: Eur. Comm. Inst. Adv. Sustain. Stud. (IASS), [126] J. Quentin, Ausbausituation der Windenergie an Land im Jahr 2020, Berlin, 2020. Potsdam, 2020. [127] K. Witsch, Trotz Milliardeninvestments und grüner Finanzprodukte: Ausbau der [93] R.A. Rodríguez, S. Becker, G.B. Andresen, D. Heide, M. Greiner, Transmission Windkraft stockt, Handelsblatt, 2021. needs across a fully renewable european power system, Renew. Energy 63 (2014) [128] O. Renn, I. Wolf, D. Setton, Soziales Nachhaltigkeitsbarometer der Energiewende, 467–476, https://doi.org/10.1016/j.renene.2013.10.005. 2020, https://doi.org/10.7802/2120. [94] German Advisory Council on the Environment, Pathways Towards a 100% [129] A. Gardt, T. Broekel, P. Gareis, Blowing against the winds of change? The Renewable Electricity System, 2011. relationship between anti-wind initiatives and wind turbines in Germany, in: Pap. 15 D. Süsser et al. Energy Research & Social Science 92 (2022) 102775 Evol. Econ. Geogr. (PEEG), Utr. Univ. Dep. Hum. Geogr. Spat. Planning, Gr. Econ. A. Popp, M. van den Berg, D. van Vuuren, E.G. Hertwich, Environmental co- Geogr, 2021, p. 2119. benefits and adverse side-effects of alternative power sector decarbonization [130] V. Kati, C. Kassara, Z. Vrontisi, A. Moustakas, The biodiversity-wind energy-land strategies, Nat. Commun. 10 (2019) 1–13, https://doi.org/10.1038/s41467-019- use nexus in a global biodiversity hotspot, Sci. Total Environ. 768 (2021), 13067-8. 144471, https://doi.org/10.1016/j.scitotenv.2020.144471. [150] T. Gibon, R. Wood, A. Arvesen, J.D. Bergesen, S. Suh, E.G. Hertwich, [131] D.P. Vasilakis, D.P. Whitfield, S. Schindler, K.S. Poirazidis, V. Kati, Reconciling A methodology for integrated, multiregional life cycle assessment scenarios under endangered species conservation with wind farm development: cinereous vultures large-scale technological change, Environ. Sci. Technol. 49 (2015) 11218–11226, (Aegypius monachus) in South-Eastern Europe, Biol. Conserv. 196 (2016) 10–17, https://doi.org/10.1021/acs.est.5b01558. https://doi.org/10.1016/j.biocon.2016.01.014. [151] C. Mutel, B. Cox, Wurst (n.d.), https://github.com/polca/wurst. [132] Bund-Länder-Kooperationsausschuss, Bericht des Bund-Länder- [152] I. Capellán-Pérez, I. de Blas, J. Nieto, C. de Castro, L.J. Miguel, Ó. Carpintero, Kooperationsausschusses zum Stand des Ausbaus der erneuerbaren Energien M. Mediavilla, L.F. Lobejón, N. Ferreras-Alonso, P. Rodrigo, F. Frechoso, sowie zu Flächen, Planungen und Genehmigungen für die Windenergienutzung an D. Álvarez-Antelo, MEDEAS: a new modeling framework integrating global Land, 2021. biophysical and socioeconomic constraints, Energy Environ. Sci. 13 (2020) [133] T. Tröndle, Supply-side options to reduce land requirements of fully renewable 986–1017, https://doi.org/10.1039/c9ee02627d. electricity in Europe, PLoS One 15 (2020) 1–19, https://doi.org/10.1371/journal. [153] N. Martin, L. Talens-Peiró, G. Villalba-Méndez, R. Nebot-Medina, C. Madrid- pone.0236958. López, The Energy Transition Beyond Climate Neutrality: Deeper Evaluations of [134] Federal Ministry for the Economy and Climate Action (BMWK), Kerninhalte der Impacts and Constraints in Future Systems, 2022. Manuscript. Referentenentwürfe des BMWK zur Novelle des Novelle des Wind-auf-See- [154] European Commission, Study on the EU’s list of Critical Raw Materials (2020). Gesetzes und zum EEG-Entlastungsgesetz. https://www.bmwi.de/Redaktion/DE Factsheets on Non-Critical Raw Materials, Brussels, Belgium, 2020, https://doi. /Downloads/E/Kerninhalte_EEG_WindSeeG.pdf?__blob=publicationFile&v=4, org/10.2873/867993. 2022. [157] D. Süsser, H.Al Rakouki, J. Lilliestam, The QTDIAN modelling [135] B.W. (BWE), Windenergie in Deutschland - Zahlen und Fakten. https://www. toolbox–quantification of social drivers and constraints of the diffusion of energy wind-energie.de/themen/zahlen-und-fakten/deutschland/, 2021. technologies, in: Deliverable 2.3. Sustainable Energy Transitions Laboratory [136] P. Anagnostopoulos, N.A. Spyridaki, A. Flamos, A “new-deal” for the development (SENTINEL) Project, 2021, https://doi.org/10.48481/iass.2021.015. of photovoltaic investments in Greece? A parametric techno-economic [158] B. Best, J. Thema, C. Zell-Ziegler, F. Wiese, J. Barth, S. Breidenbach, assessment, Energies 10 (2017) 1173, https://doi.org/10.3390/en10081173. L. Nascimento, H. Wilke, Building a database for energy sufficiency policies, [137] R. Loulou, M. Labriet, ETSAP-TIAM: the TIMES integrated assessment model part F1000Research 11 (2022) 229, https://doi.org/10.12688/ I: model structure, Comput. Manag. Sci. 5 (2008) 7–40, https://doi.org/10.1007/ f1000research.108822.1. s10287-007-0046-z. [159] M. Degel, M. Christ, L. Becker, J. Grünert, Sozial-ökologische und technisch- [138] Environmental E.S.A. A. Center for Energy, in: Energy and Power Evaluation ökonomische Modellierung von Entwicklungspfaden der Energiewende, Berlin, Program (ENPEP-BALANCE): Brief Model Overview - Version 2.25, 2008, pp. 1–7. 2016. [139] J. Santisirisomboon, B. Limmeechokchai, S. Chungpaibulpatana, Impacts of [160] F. Lombardi, B. Pickering, E. Colombo, S. Pfenninger, Policy decision support for biomass power generation and CO taxation on electricity generation expansion renewables deployment through spatially explicit practically optimal alternatives, planning and environmental emissions, Energy Policy 29 (2001) 975–985, Joule 4 (2020) 2185–2207, https://doi.org/10.1016/j.joule.2020.08.002. https://doi.org/10.1016/S0301-4215(01)00028-3. [161] R. McKenna, J.M. Weinand, I. Mulalic, S. Petrović, K. Mainzer, T. Preis, H.S. Moat, [140] Ministry of Environment and Energy, 1st National Renewable Energy Action Plan Scenicness assessment of onshore wind sites with geotagged photographs and (NREAP), 2010. impacts on approval and cost-efficiency, Nat. Energy 6 (2021) 663–672, https:// [141] S. Michas, V. Stavrakas, S. Papadelis, A. Flamos, A transdisciplinary modeling doi.org/10.1038/s41560-021-00842-5. framework for the participatory design of dynamic adaptive policy pathways, [162] T. Zhang, W.J. Nuttall, Evaluating government’s policies on promoting smart Energy Policy 139 (2020), https://doi.org/10.1016/j.enpol.2020.111350. metering diffusion in retail electricity markets via agent-based simulation, [142] I. Koumparou, G.C. Christoforidis, V. Efthymiou, G.K. Papagiannis, G. J. Prod. Innov. Manag. 28 (2011) 169–186, https://doi.org/10.1111/j.1540- E. Georghiou, Configuring residential PV net-metering policies – a focus on the 5885.2011.00790.x. Mediterranean region, Renew. Energy 113 (2017) 795–812, https://doi.org/ [163] M. Melliger, E. Chappin, Phasing out support schemes for renewables in 10.1016/j.renene.2017.06.051. neighbouring countries: an agent-based model with investment preferences, Appl. [143] S. Papadelis, V. Stavrakas, A. Flamos, What do capacity deployment rates tell us Energy 305 (2022), 117959, https://doi.org/10.1016/j.apenergy.2021.117959. about the efficiency of electricity generation from renewable energy sources [164] M. Melliger, J. Lilliestam, Effects of coordinating support policy changes on support measures in Greece? Energies 9 (2016) https://doi.org/10.3390/ renewable power investor choices in Europe, Energy Policy 148 (2021), 111993, en9010038. https://doi.org/10.1016/j.enpol.2020.111993. [144] A. Flamos, A sectoral micro-economic approach to scenario selection and [165] D. Süsser, A. Ceglarz, H. Gaschnig, V. Stavrakas, A. Flamos, G. Giannakidis, development: the case of the Greek power sector, Energies 9 (2016) 1–12, https:// J. Lilliestam, Model-based policymaking or policy-based modelling? How energy doi.org/10.3390/en9020077. models and energy policy interact, Energy Res. Soc. Sci. 75 (2021), 101984, [145] S. Psomas, in: Status and Outlook of the Greek PV Market, Hell. Assoc. Photovolt. https://doi.org/10.1016/j.erss.2021.101984. Co, 2018, pp. 1–14. [166] C. Mutel, X. Liao, L. Patouillard, J. Bare, P. Fantke, R. Frischknecht, M. Hauschild, [146] S. Tselepis, The PV market developments in Greece, net-metering study cases, in: O. Jolliet, D. Maia de Souza, A. Laurent, S. Pfister, F. Verones, Overview and Conf. 31st EUPVSEC, Hamburg, 2015. recommendations for regionalized life cycle impact assessment, Int. J. Life Cycle [147] D. Setton, Soziales Nachhaltigkeitsbarometer der Energiewende 2018, 2019, Assess. 24 (2019) 856–865, https://doi.org/10.1007/s11367-018-1539-4. https://doi.org/10.2312/iass.2019.002. [167] D. Süsser, A. Ceglarz, H. Gaschnig, V. Stavrakas, G. Giannakidis, A. Flamos, [148] M. Pehl, A. Arvesen, F. Humpenöder, A. Popp, E.G. Hertwich, G. Luderer, A. Sander, J. Lilliestam, The use of energy modelling results for policymaking in Understanding future emissions from low-carbon power systems by integration of the EU. Deliverable 1.1, in: Sustainable Energy Transitions Laboratory life-cycle assessment and integrated energy modelling, Nat. Energy 2 (2017) (SENTINEL) Project, European Commission, 2020. 939–945, https://doi.org/10.1038/s41560-017-0032-9. [168] D. Süsser, H. Gaschnig, A. Ceglarz, V. Stavrakas, G. Giannakidis, A. Flamos, [149] G. Luderer, M. Pehl, A. Arvesen, T. Gibon, B.L. Bodirsky, H.S. de Boer, O. Fricko, A. Sander, J. Lilliestam, Models for the European Energy Transition: Your M. Hejazi, F. Humpenöder, G. Iyer, S. Mima, I. Mouratiadou, R.C. Pietzcker, Questions, Your Needs!, Workshop Synthesis Report, 2020. 16