Decision support models in climate policy

European Journal of Operational Research 280 (2020) 1–24

Contents lists available at ScienceDirect

European Journal of Operational Research journal homepage: www.elsevier.com/locate/ejor

Invited Review

Decision support models in climate policy Haris Doukas∗, Alexandros Nikas Decision Support Systems Laboratory, School of Electrical and Computer Engineering, National Technical University of Athens, Iroon Politechniou 9, 157 80, Athens, Greece

a r t i c l e

i n f o

Article history: Received 29 April 2017 Accepted 7 January 2019 Available online 12 January 2019 Keywords: Decision support Climate policy Fuzzy cognitive maps Multiple criteria decision making Portfolio analysis

a b s t r a c t Climate change is considered among the most critical risks that global society faces in this century. So far, climate policy strategies have been evaluated by means of a variety of climateeconomy models, or Integrated Assessment Models (IAMs), in the aim of supporting climate-related decision making. However, their inherent complexity, the number and nature of driving assumptions, and usual exclusion of stakeholders from the modelling process raise the issue of the extent to which they can provide fruitful insights for policy makers. Moreover, as with all modelling frameworks, IAMs inevitably fail to incorporate all relevant types of uncertainty and risk when used as stand-alone tools. This exclusion can have a significant impact on the model outcomes, but can be mitigated if experts’ knowledge is elicited in a structured manner and effectively taken into account, towards identifying such factors or reducing respective knowledge gaps. At the same time, a growing number of research publications have been suggesting decision support frameworks for assessing specific aspects in climate policy, based on “bottom-up” approaches and participatory processes. The objective of this paper is to provide a critical review of such frameworks—namely Portfolio Analysis, Multiple Criteria Decision Making and Fuzzy Cognitive Maps—in order to explore their strengths and weaknesses in this area, and propose a new integrative approach, appropriately exploiting blends of these frameworks, to productively complement IAMs, towards enhancing climate policy support. © 2019 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license. (http://creativecommons.org/licenses/by-nc-nd/4.0/)

1. Introduction As the balance of scientific evidence suggests (Benestad et al., 2016; Soderholm, 2007), anthropogenic greenhouse gas (GHG) emissions emanating from economic activities significantly impact the environment and amplify climate change, with the use of energy for electricity and heat production being the key source of these emissions (Edenhofer et al., 2014). This, in turn, inarguably poses one of the largest sustainability threats of our time and affects not only the environment but various, if not all, sectors of the economy, both directly and indirectly. Motivated by the need to study the dynamic interactions between the economy, energy and climate dimensions, Integrated Assessment Models (IAMs) have long been used to support climate policy making. Although their contribution to this challenging task is uncontested, the research community has long been questioning the extent to which they have actually supported policy makers (e.g. Kelly & Kolstad, 1999; Schneider, 1997; Watkiss, Downing, & Dyszynski, 2010). ∗

Corresponding author. E-mail addresses: [email protected], [email protected] (H. Doukas).

Climate change as well as mitigation and adaptation policy feature significant levels and different types of uncertainties and risks. These regard issues ranging from future carbon accumulation levels to socio-economic developments that may promote or hinder the adoption and implementation of policy measures in this direction, such as the capacity to fund technological innovation or the levels of societal acceptance of policies. Uncertainty can be framed as a broad concept that refers to a general lack of knowledge or agreement upon possible outcomes and their probabilities. Risk can thus be defined as a negative possible outcome that stems from an uncertainty and largely depends on the focus of the study, regardless of whether it can be accurately quantified as a probability or be attributed a qualitative likelihood, e.g. based on stakeholders’ experience. From a scientific perspective and as with all modelling frameworks, IAMs inevitably fail to incorporate all relevant types of uncertainty and risks when used as stand-alone tools. This exclusion can have a significant impact on the model outcomes, but can be mitigated if experts’ knowledge, which can prove valuable in identifying such factors or reducing respective knowledge gaps, is elicited in a structured manner and effectively taken into account. At the same time, formalised modelling

https://doi.org/10.1016/j.ejor.2019.01.017 0377-2217/© 2019 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license. (http://creativecommons.org/licenses/by-nc-nd/4.0/)

2

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

frameworks may impose representations and restrictions not justified by the underlying knowledge, may aggregate results, and cannot directly represent all policies (Agrawala et al., 2010), for instance soft measures aimed at human behaviour (e.g. dissemination of knowledge through education and public information campaigns). This is an expected weakness of climate-economy modelling frameworks, which can only indirectly represent and calculate the economic impacts of climate policies. From a policy making perspective, climate policy makers usually find it hard to understand the complex processes of the very large number of existing IAMs, keep track of the various assumptions driving their modelling simulations, and trust their black-box nature (Kelly & Kolstad, 1999). Furthermore, as with every other stakeholder group, policy makers are not always actively involved in the modelling processes and are only perhaps engaged in some limited, preliminary discussions towards formulating parts of the assumptions used (van Vliet, Kok, & Veldkamp, 2010). These weaknesses of climate-economy models require the adoption of additional frameworks for supporting climate policy design. These should include frameworks and tools that not only bring policy makers and other stakeholders closer to the process, but also enable the mobilisation of their tacit knowledge and experience towards taking into account various implementation and consequential risks and uncertainties. Among these, there can be found different levels of detail, capacity to identify and assess risks and uncertainties, and flexibility to include stakeholders, cover multiple sectors and customise the scope of study. These include Cost-Benefit Analysis (CBA), Fuzzy Cognitive Maps (FCMs), Life Cycle Assessment (LCA), Multiple Criteria Decision Making (MCDM), Portfolio Analysis (PA), System Mapping (SM), and Systems of Innovation frameworks, e.g. Technological Innovation Systems (TIS) and Multi-Level Perspective (MLP). CBA has been established as a decision support tool for assessing the economic efficiency of interventions in multiple domains, including climate policy (e.g. Tol, 2012; van den Bergh, 2004; van den Bergh & Boltzen, 2015); however, it appears to feature significant limitations regarding its capacity to incorporate uncertainty or to quantify non-market goods and social and environmental distributional impacts, as outlined by Shreve and Kelman (2014). LCA and decision analysis have a long-established connection (Miettinen & Hämäläinen, 1997) but, with regard to climate policy, dependence on data availability, poor definition of boundaries, time requirements and assumptions may limit its potential. Nevertheless, LCA has been primarily used in environmental impact assessment studies, with climate-related implications for the power (e.g. Georgakellos, 2012; Stamford & Azapagic, 2014), transport (e.g. Ashnani, Miremadi, Johari, & Danekar, 2015; Di Lullo, Zhang, & Kumar, 2016), building (e.g. Le Téno & Mareschal, 1998; Tsai, Yang, Chang, & Lee, 2014), and agriculture and land use (e.g. Humpenöder, Schaldach, Cikovani, & Schebek, 2013; Røyne, Penaloza, Sandin, Berlin, & Svanström, 2016) sectors. SM is a purely qualitative, stakeholder-driven tool that features relative flexibility to look into various aspects of a system and its links to policy, and can help identify policy-related risks, but has only recently been framed in the climate policy domain (Nikas et al., 2017). Both TIS and MLP focus on technological innovation and are, therefore, underexploited in this domain; they are employed primarily for socio-technical studies in the power (e.g. Edsand, 2017; Moallemi, de Haan, Webb, George, & Aye, 2017) and transport (Auvinen, Ruutu, Tuominen, Ahlqvist, & Oksanen, 2015; Hellsmark & Jacobsson, 2009) sectors. Despite their capacity to identify barriers and risks, however, both frameworks are strictly qualitative and appear to provide little insight into long-term policy (Lai, Ye, Xu, Holmes, & Lambright, 2012) and scale coverage (Geels, 2012), which are indeed important parameters in climate policy. In this study, we distinguish FCMs, MCDM, and PA, which a growing number of re-

search publications seem to suggest as promising routes to bringing policy makers (and stakeholders in general) closer to the modelling processes, in response to the highlighted weakness of IAMs, as well as to supporting decision making in various climate policyrelated fields and application areas. Acknowledging that these decision support frameworks are, compared to IAMs, significantly less detailed but can add value in other dimensions of climate policy assessment, the aim of this paper is to provide a detailed critical review of these frameworks and to explore their capacity to support decision making in specific climate policy problems. In addition, we propose a truly integrative scientific approach, appropriately exploiting blends of the three frameworks, to productively complement IAMs, towards enhancing climate policy support. To this end, the following section provides a concise overview of the various types of IAMs, while Sections 3– 5 present a detailed review of FCMs, MCDM and PA respectively. Finally, Section 6 features a discussion on the findings of our review, regarding the various manners in which these frameworks can potentially further contribute to climate policy, and strengthens this suggestion by means of a real-world climate policy problem. Section 7 summarises the conclusions of this study. It should be noted that the publications reviewed for the purposes of this paper were identified primarily on the basis of searches on the Scopus and Google Scholar academic search engines, for various keywords (e.g. “climate policy” AND “multiple criteria”, “climate change” AND “multicriteria”, “mitigation” AND “MCDM”, etc. for MCDM), and included scientific journal articles, book chapters and conference papers. Furthermore, broader searches were performed in specific journals and books, after reviewing the number and thematic focus of the associated papers retrieved in the initial results. 2. An overview of Integrated Assessment Models and the need for integration The great variety of IAMs to some extent reflects the range of underlying scientific fields that influence the design of their structure as well as the sets of assumptions driving their simulations, and of the different questions they help to address. There exist a number of detailed reviews of quantitative models employed in the climate policy literature, each one focusing on different aspects of these models. For instance, Kelly and Kolstad (1999), at a very early stage, and Stanton, Ackerman, and Kartha (2009), much later, delved into the different aspects of IAMs: the endogenous and exogenous specification of technological change, the intentional exclusion of equity across time and space from most models, and the scientific uncertainties associated with the projections of future climate outcomes and resulting damages; Füssel (2010) emphasised the aspect of adaptation to climate change and how that is incorporated in IAMs; Ortiz and Markandya (2010) focused on the mathematics of climate-economy models and, in particular, on the damage functions they use; while Schwanitz (2013) more recently proposed a methodological framework for evaluating the different climate-economy models. Other more detailed or narrowly-focused reviews can be found in the literature (e.g. Dowlatabadi, 1995; Parson & Fisher-Vaden, 1997; Rana & Morita, 20 0 0; Soderholm, 20 07; Wei, Mi, & Huang, 2015); the large number of such reviews and the major differences among them are indicators of two factors: diversity in and significant complexity of climate-economy modelling frameworks. First, it is both natural and desired that, in such a broad subject that can be analysed in so many different perspectives with distinct methodological approaches, diversity exists; however, this proliferation of IAMs can make it very hard to understand the strengths and weaknesses of each model and appreciate which is appropriate for each application. Secondly, it is expected that modelling frameworks that seek to represent the daunting,

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

multidisciplinary interconnection between climate, economy, energy and environment are themselves quite complex. This complexity shows that their outcomes are results of thoroughly elaborated processes but, on the other hand, can make it very hard for researchers and scientists, let alone policy makers, to understand their attributes and thus trust the outputs and feel confident about the results. The primary objective of this section is to provide a quick but concrete overview and categorisation of the various IAMs that will be as consistent as possible, to enable drawing conclusions with regard to their effectiveness in actively supporting the challenging process of climate policy making. This section also aims to highlight key weaknesses of IAMs, the need to address which shapes the integration potential of these models with other decision support frameworks. The various classifications found in the literature do not necessarily align with each other, since they are drawn based on different criteria. Certain studies (e.g. Füssel, 2010) classify IAMS according to the analytical frameworks they are applied to, some (e.g. Stanton et al., 2009) according to their structure, and others (e.g. Ortiz & Markandya, 2010) based on how they use and integrate their fundamental components (climate, energy, impacts) with economy. Drawing from these classifications as well as the climate-economy modelling literature, six different families of IAMs can be distinguished and are presented: welfare optimisation, general equilibrium, partial equilibrium, energy system, macroeconometric, and other models. It is worth pointing out, however, that the nature of these models does not allow for a completely consistent taxonomy, in a sense that certain IAMs can inevitably fit in more than one categories. For example, the AIM/Dynamic Global (Masui et al., 2006) and RICE (Nordhaus, 1994) welfare optimisation models can also be considered as general equilibrium models. Neoclassical or optimal growth or welfare optimisation models focus on maximising the discounted present value of welfare (utility) across all time periods, by choosing how many emissions to abate in each time period. These IAMs, therefore, attempt to find the balance between present and future consumption towards maximising overall welfare. The key feature that transforms these models into climate-economy models is the integration of the natural capital of the climate (Nordhaus, 2008), which is depleted by GHG emissions and increased by investments in abatement. Apart from their relative transparency, the strengths and weaknesses of the models of this category generally depend on their individual structure and methodological approach. For example, the MERGE model (Manne & Richels, 2005) covers carbon emissions for both current and future technologies but the co-existence of non-market damages induces high uncertainties and its large size increases the model’s complexity; the GRAPE model (Kurosawa, Yagita, Zhou, Tokimatsu, & Yanagisawa, 1999) assesses carbon dioxide storage and covers emissions from land use but at the same time excludes methane and features great uncertainty regarding carbon storage; while FEEM-RICE (Buonanno, Carraro, & Galeotti, 2003) features consistency in large parameter changes but does not consider the growing effectiveness of carbon sequestration technologies. Computable general equilibrium (CGE) models represent the economy as a set of strongly intertwined economic sectors. They aim at determining a set of prices that simultaneously equate supply and demand in each sector, with consumers maximising utility, producers maximising profit and prices being the driving factor of equilibrium. These models can support climate policy making by measuring how “shocks” on the economy, i.e. changes to one or more exogenous parameters, can affect market equilibrium, thus providing insight into the market behaviour against a specific policy or sets of policies. This way, CGEs can determine the profit or loss of welfare to the representative agent resulting from productivity shocks, but do not capture impacts that are not closely linked to the market, such as losses in biodiversity or certain so-

3

cietal risks. Again, there are significant differences among models of this category. For instance, the WORLDSCAN model (Lejour, Veenendaal, Verweij, & van Leeuwen, 2006) can be used both for the construction of long-term scenarios and as an instrument for multilateral policy impact assessments (e.g. Bollen & Gielen, 1999) but excludes numerous emissions; while ICES (Bosello, De Cian, Eboli, & Parrado, 2009) covers all Kyoto GHG emissions and offers flexibility in regional or sectorial aggregations (e.g. Parrado & De Cian, 2014) but only accounts for primary resources for driving economic growth. Partial equilibrium models, on the other hand, are based on the same theoretical foundations as general equilibrium models but differ from them by focusing on a specific sector or number of sectors, assuming that conditions in the remaining economy remain constant. This approach is usually justified when changes affecting the sector or sectors under examination are expected to have insignificant impacts on the remaining economy. As a result, they feature a finer level of detail regarding the focus sectors but cannot capture the full implications of the propagation of sectorial changes across all economy sectors, nor do they usually attempt to cover all major impacts of climate change, as is the case with the MINICAM model (Edmonds & Reiley, 1985). While partial equilibrium models focus on estimating damages of changes associated with climate change, within a particular sector, energy system models are exclusively bottom-up tools that focus on the key sector determining greenhouse gas emissions and costs arising from mitigation or adaptation policies, i.e. the power sector: according to IPCC (Edenhofer et al., 2014), the burning of coal, natural gas, and oil for electricity and heat is the largest single source of global greenhouse gas emissions. Energy system models are in essence disaggregated representations of the energyeconomy system with a clear focus on existing and emerging energy technologies. These can simulate alternative energy futures (Mundaca, Neij, Worrell, & McNeil, 2010) in the aim of covering aspects, not limited to achieving emission targets with the lowest costs but also including determining technological opportunities or costs of alternative policies. Some taxonomies further classify energy system models into simulation and optimisation models (e.g. Jebaraj & Iniyan, 2006; or Worrell, Ramesohl, & Boyd, 2004). Very popular models of this family include the MARKAL (Fishbone & Abiock, 1981) and MESSAGE (Messner, 1997) models. Despite the level of detail that is associated with this highly disaggregated approach—which can sometimes lead to large complexity, as in the GENIE model (Mattsson & Wene, 1997), or exclusion of other sectors or even of a climate module, as in the DNE21+ model (Sano, Akimoto, Homma, & Tomoda, 2006)—they can sometimes ignore niche markets and learning aspects, such as the GET-LFL model (Hedenus, Azar, & Lindgren, 2006). Far from this distinction, an interesting family of IAMs is that of macroeconometric models. These are hybrid models that tend to integrate top-down macro models with bottom-up energy system modules, towards providing a non-linear picture of economic change in a cross-sectoral framework, by tracking the time path of economy through short-run disequilibrium adjustments. Transactions among economic sectors are described by input-output models and historical-data-driven damage functions allow for forecasting responses to climate policies. One of their major differences from and strengths against other modelling frameworks is that they are disequilibrium models with demand and supply approaching equilibrium in the long run, unlike welfare optimisation and equilibrium models that assume that markets clear in the shorter or medium term. Most importantly, however, and despite recent suggestions that macroeconometric and CGE models do not significantly differ from each other (Kratena & Streicher, 2009), this family of models attempts to represent climate-economy interactions in a more realistic manner, by not attributing optimising

4

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

behaviour to the system agents. Representative models of this family include the Energy-Environment-Economy Model for Europe or E3ME (Barker & Zagame, 1995), along with its global extension, the E3MG model (Barker, Pan, Köhler, Warren, & Winne, 2006). Finally, a separate “other” model category is proposed. IAMs of this category significantly differ both from models of other categories and from one another; one thing, however, that they have in common is that they are more policy-oriented than models of other categories, whether they systematically evaluate impacts and interrelations of different policies (simulation models) or determine the optimal set of choices towards achieving a specific goal (optimisation models). For example, FUND (Tol, 1997) is a non-CGE policy optimisation model that does not evaluate the impacts of a specific policy but rather supports policy makers in understanding what optimal policies look like for a given system; while ICAM3 (Dowlatabadi, 1998) is designed to assess the mitigation cost of technological change (e.g. Dowlatabadi, 20 0 0). A rather popular yet relatively recent IAM that falls into this category is the PAGE2002 model (Hope, 2006), the top-down model that the Stern review (Stern, 2007) relied on for many of the so called aggregate climate change damages. PAGE2002 considers large-scale climate instabilities and assesses all damage function parameters through probability distributions instead of single best estimates; on the other hand, it includes only direct costs of preventing emissions and does not account for smaller benefits of GHG abatement, such as the double dividend of carbon tax (Carraro, Galeotti, & Gallo, 1996; Goulder, 1995). Aside from the large differences among the categories of IAMs, all of these models are inevitably driven by a large number of assumptions (Watkiss et al., 2010) and tend to be very complex for policy makers to actually understand and exploit their results (Nikas & Doukas, 2016). Other associated weaknesses can be summarised in their difficulty to model the impact of all possible mitigation and adaptation policy instruments or incorporate risks and uncertainties inherent in climate change and policy, and the limited extent to which they allow for the participation of stakeholders in the modelling process. The latter is particularly important, since stakeholders represent the institutional and socio-economic factors that must lead the desired technological and regulatory innovations as well as energy consumption lifestyles and patterns assumed in mitigation and adaptation policy. As such, they must belong in the heart of scientific analyses that essentially inform policy making processes. This is especially true for policy makers, who constitute a key stakeholder group and who are asked to translate the results of complex modelling processes, in which they barely participate, into policy. In pursuit of a new scientific paradigm that will bring policy makers (and stakeholders in general) closer to the modelling processes as well as improve the understanding and assessment of uncertainty (Doukas, Nikas, González-Eguino, Arto, & AngerKraavi, 2018), the following sections investigate decision support frameworks that can enhance scientific processes in support of climate policy making, in response to the highlighted weaknesses of climate-economy models. 3. A semi-quantitative modelling approach: Fuzzy Cognitive Maps Fuzzy Cognitive Mapping, as a modelling approach, has its roots in Cognitive Mapping. The latter is a purely qualitative modelling technique that aims to capture people’s values, beliefs and assumptions with regard to a particular issue, problem or domain in diagrammatic format (Eden & Ackermann, 2013), allowing for ad hoc structure (Brown, 1992) and, therefore, unconstrained freedom. A cognitive map can be defined as the graphical representation of a system, in which every node represents a particular concept within

the system and every arc represents the perceived interconnections between these concepts. Cognitive maps work as a transitional object applied by experts in order to express and understand their knowledge with regard to the structure of a particular problem domain, and can be used, inter alia, for the purpose of exploring and assessing influence, causality and system dynamics (Huff, 1990). Kosko, however, suggested that cognitive maps are too binding for knowledge-base building because causal influence between concepts involves fuzziness and cannot be adequately described by just using arcs (Kosko, 1986). He, thus, introduced the notion of FCMs, in which causal relations are also quantified by means of causal weights. By allowing feedback loops in the model, FCMs are considered to be similar to other quantitative causal models, such as system dynamics; as a semi-quantitative model, however, they do not require historical data or translation of qualitative input into quantitative systems (Jetter & Schweinfort, 2011). Fuzzy Cognitive Mapping usually involves an intensive stakeholder engagement process, during which stakeholder input is translated into concept nodes (e.g. events, policy-defined goals, system trends, transition drivers, risks and uncertainties) and causal links (i.e. weighted interrelations), which together comprise the FCM representation model. After the FCM has been designed, the systematic causal propagation (Kosko, 1986) can be analytically traced through simulations (Papageorgiou & Kontogianni, 2012), by borrowing techniques from neural networks. To this end, a simulation driver function (or activation function) and a transfer function (also known as threshold or transformation function) are employed in order to allow for this causal propagation to take place and be captured. These simulations may converge to a fixed point, a limit cycle (loop), a limit torus, or a chaotic attractor in the fuzzy cube (Dickerson & Kosko, 1994), depending on three dimensions: the FCM structure, the link weights and the initial state vector. This analysis enables the system under examination to be stresstested against multiple what-if scenarios, by performing simulations while changing one of the three dimensions and keeping the other two dimensions fixed. Comparisons between the results of different what-if scenarios can, therefore, be used to support decision making or scenario building and analysis (Stach, Kurgan, & Pedrycz, 2010). As a semi- or quasi-quantitative modelling technique, FCMs offer various advantages compared to both other participatory research methods and more strictly-structured modelling frameworks. Given their high flexibility and unconstrained freedom, their little if any dependence on data availability and the fact that they are built on and driven by human expertise and knowledge, FCMs have gained considerable research interest during the last decade. Especially with regard to policy making, FCMs are known to bring modellers and experts (policy makers included) together, making the latter feel greatly appreciated, as an essential part of an otherwise challenging process that is policy making, and thus more willing to trust and make use of their outcomes. As a result, fuzzy cognitive mapping has been widely applied in the diverse and multidisciplinary domains of environmental and energy policy. Explicitly for the purposes of climate policy support, there appear to be only a small number of fuzzy cognitive mapping applications. In a broader perspective, however, there are many studies revolving around the problem domain of climate change and having significant direct or indirect implications to climate policy making. These feature a rather diverse set of application areas, mainly comprising climate change scenario analyses and policy strategy evaluation, for the purposes of supporting environmental, climate change mitigation and climate change adaptation policy and decision making. Scenario analysis applications mainly study alternative climate change scenarios or alternative future system developments against these scenarios (e.g. Anezakis, Dermetzis, Iliadis, & Spartalis, 2016), without necessarily targeting specific policies;

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

5

Table 1 Overview of climate policy related studies in the FCM literature, with regard to their scope, sector and means of engagement (I=Interviews, S=Survey/Questionnaire, W=Workshop, O=Other). Study

Scope of study

Focus sector Agriculture

Amer et al. (2011) Amer et al. (2016) Anezakis et al. (2016) Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Ceccato (2012) Çelik et al. (2005) Christen et al. (2015) Ghaderi et al. (2012) Giordano et al. (2017) Gray et al. (2013) Gray et al. (2014) Gray et al. (2015) Hobbs et al. (2002) Hsueh (2015) Huang et al. (2013) Jetter and Schweinfort (2011) Kafetzis et al. (2010) Karavas et al. (2015) Kayikci and Stix (2014) Kontogianni et al. (2012) Kontogianni et al. (2013) Kottas et al. (2006) Kyriakarakos et al. (2012) Kyriakarakos et al. (2014) Lopolito et al. (2011) Mallampalli et al. (2016) Meliadou et al. (2012) Mourhir et al. (2016) Natarajan et al. (2016) Nikas and Doukas (2016) Olazabal and Pascual (2016) Ortolani et al. (2010) Özesmi and Özesmi (2003) Özesmi (2006a) Özesmi (2006b) Papageorgiou and Kontogianni (2012) Papageorgiou et al. (2011) Peng et al. (2016) Rajaram and Das (2010) Reckien (2014) Sacchelli (2014) Samarasinghe and Strickert (2013) Shiau and Liu (2013) Singh and Nair (2014) Soler et al. (2012) van Vliet et al. (2010) Vanwindekens et al. (2013) Vassilides and Jensen (2016) Wildenberg et al. (2010) Zhang et al. (2013) Zhao et al. (2014)

Scenario analysis Scenario analysis Scenario analysis Scenario analysis Scenario analysis Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Scenario analysis; Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Scenario analysis Policy evaluation Scenario analysis Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Scenario analysis Policy evaluation Policy evaluation Scenario analysis; Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation Policy evaluation

environmental policy covers ecosystem conservation and environmental decision making (e.g. Özesmi & Özesmi, 2003; Vassilides & Jensen, 2016); climate change adaptation policy refers to FCM applications that explicitly study system resilience and evaluate actions in the aim of responding to climate change (Gray et al., 2014; Reckien, 2014); while all other applications revolve around policy choices towards mitigation mainly with regard to the development of agriculture and land use (e.g. Singh & Nair, 2014), renewable energy sources (e.g. Hsueh, 2015) and electricity planning (e.g. Karavas, Kyriakarakos, Arvanitis, & Papadakis, 2015), and the transition of the transportation sector (Kontogianni, Tourkolias, & Papageorgiou, 2013; Shiaua & Liu, 2013). In order to provide a consistent and collectively exhaustive classification, these are divided into economic sectors (Table 1).

Means of engagement Environment

Power √ √

Transport

√ √ √ √ √ √ √ √ √ √ √ √ √

√ √ √

√ √ √ √ √ √ √ √ √

√ √ √ √ √

√ √

√ √ √ √ √

√ √ √ √

√ √ √

√ √

√

√ √ √ √ √

√ √ √ √ √ √ √

I, S W O S, O S, I I, S, W I I, W S I I, W I, S, O W I I, O O I I O S I I O O I I S I W, S S, O I I, S, O S, I I I S, I I O O S I I, S I W I I, O W I, S I I, W I, S, O S, I

The approach employed for eliciting stakeholder input and translating it into visual representation of the problem domain is of great importance; different stakeholder engagement approaches allow for different levels of detail to be incorporated into the model. As a first step, experts are either asked to directly design their personal FCMs during interviews, in groups or by means of surveys, or are facilitated into collectively designing a unique fuzzy cognitive map. The most commonly employed approach of faceto-face interviews is followed by workshops and semi-structured questionnaires, although there have been instances in the literature where modellers built their model based on historical data only. Table 1 provides an overview of all FCM applications with direct or indirect climate policy implications, along with the selected participatory approaches.

6

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Table 1 suggests that the vast majority of FCM studies in the literature reviewed have been carried out to support the design of environmental policy and planning, while less studies appear to have explicit implications for climate action and policy. This can be attributed to the geographic level and scope of climate policy making, which is mostly addressed at the regional or global level, as opposed to environmental modelling and scenario analysis, which are most often studied at the local level. After the map has been constructed and using simulation techniques from artificial neural networks, the model is simulated in order to allow for causal propagation to take place and examine the attitude of the system, given its unique combination of structure, weights and initial state vector. It should be noted however that, due to the need to respond to climatic change in time (adaptation) and/or to reduce greenhouse gas emissions (mitigation) in strictly defined pathways, the time dimension must also be taken into account when designing climate policy, an aspect that traditional FCM approaches fail to address (van Vliet et al., 2010). Only a small number of studies modify the original activation function in an effort to incorporate the notion of time into the FCM methodology. For example, Nikas and Doukas (2016) identified the need to incorporate the critical concept of time into FCM simulations, when developing robust policy strategies for climate mitigation, and therefore translated each iteration into a specific time period and assigned a time delay (or lag) to each causal relationship. Biloslavo and Dolinšek (2010) proposed a method that transforms the weight matrix into a time function. Finally, Mourhir, Rachidi, Papageorgiou, Karim, and Alaoui (2016) used a Dynamic Rule-Based FCM approach, in which weights are adapted dynamically during a simulation. Eventually, the calculated output of the model shows how the system reacts under the assumptions provided by the stakeholders. Comparisons between the final state vectors of the examined alternatives should be drawn in order to assess to what extent the desired transition has been promoted, by activating each option, policy or other driver. A specific set of concepts is selected as evaluation criteria; in policy making, the larger the value of the goal concept is at the end of the simulation, the better the selected policy appears to be considered by stakeholders. In most cases, system dynamics are evaluated by looking at the value of a specific concept, but given the multidisciplinary nature of climate policy problem domains, many researchers evaluated their systems by looking at the final state vector or a set of concepts. For example, Lopolito, Nardone, Prosperi, Sisto, and Stasi (2011) evaluated four policies regarding the development of the bio-refinery industry by using a weighted average of the values of the key development concept and a number of negative side effects. Nevertheless, no study in the literature explicitly aims to evaluate policy strategies, i.e. sets of policy instruments, but rather against individual policies. Finally, in this research area in particular, FCMs have been implemented both as a stand-alone methodology and as part of diverse methodological frameworks or integrated with other methodologies, frameworks and tools (Table 2). Drawing from Table 2, it is evident that only a few studies have integrated FCMs with quantitative models. Among these, Anezakis et al. (2016) explore alternative climatic scenarios by means of a climate model, while van Vliet et al. (2010) evaluate environmental policies; in fact, only two studies (Nikas & Doukas, 2016; Mallampalli et al., 2016) refer to links between FCMs and IAMs, but are limited to only describing the methodological framework and respective potential. 4. Multiple Criteria Decision Making and climate policy Multiple Criteria Decision Making (MCDM), sometimes referred to as Multiple Criteria Decision Aid or Analysis (MCDA) (Roy, 1990),

is a sub-discipline of operations research. It aims to support decision making in complex problems where multiple points of view (criteria) must be taken into account before reaching a meaningful solution (Govindan & Jepsen, 2016). MCDM can support various stages of decision making, including problem structuring, preference modelling, construction of criteria aggregation models, and design of interactive solution procedures (Doumpos & Zopounidis, 2011). It has been considerably evolving since it first appeared (Roy & Vanderpooten, 1997) and has been receiving increasing attention (Govindan & Jepsen, 2016), especially during the last decades (Behzadian, Kazemzadeh, Albadvi, & Aghdasi, 2010). Despite the fact that climate policy studies initially lacked the knowledge and guidelines regarding the employment of MCDM approaches (Borges & Villavicencio, 2004), they have attracted increasing attention over the last decade in studies related to climate mitigation and adaptation as well as to policy making in various economic sectors where GHG emissions reduction is among the main evaluation criteria (Scrieciu & Chalabi, 2014). This remarkable boom can be partly attributed to the growing proliferation of respective methodological frameworks; the increasing engagement, collaboration and participation of expert decision makers in modern modelling activities (Voinov & Bousquet, 2010); and the need for developing integrated methodologies for addressing problems in this domain that generally comprise numerous stages, in which MCDM can significantly contribute. Another major driver of their ever-growing use can be found in their large popularity in energy policy problems combined with the fact that decarbonising the power sector lies at the heart of climate policy (Doukas, 2013). Another factor contributing to said increase can be found in international and European Union (EU) policy efforts for sustainable development. In particular, sustainability and respective implications for the environment are intertwined with decarbonisation and climate change mitigation. At the same time, sustainable development has been framed in reference to its multiple dimensions or pillars (economy, environment, society), as outlined in the so-called Brundtland report of the World Commission on Environment and Development (Brundtland Commission, 1987); and hence has long been studied by means of multi-criteria approaches. In general, a multitude of MCDM frameworks have been employed in energy planning and sustainable development studies (Diakoulaki, Antunes, & Gomes Martins, 20 05; Munda, 20 05); a selection of applications in these two areas can be found in Part VII of Greco, Ehrgott, and Figueira (2016). When reviewing the MCDM literature and taking into account the search criteria outlined in Section 1, 73 studies were found to be relevant to the climate policy field (Table 3). In these studies, multicriteria analyses were involved either directly in policy evaluation or indirectly in other areas, in which climate mitigation or adaptation criteria were included. Most studies focused on the assessment of different technologies, closely followed by those revolving around climate policy instrument or strategy comparison and evaluation, while a limited number of researchers used multicriteria processes for the purposes of selecting projects among various alternatives or analysing scenarios with climate policy implications. Finally, only one study performed risk assessment (Branco et al., 2012), one aimed at emissions-driven EU country comparisons (Dace & Blumberga, 2016), and two focused on the identification and prioritisation of evaluation factors and indicators for the assessment of energy projects (Heo, Kim, & Boo, 2010; and Luthra, Mangla, & Kharb, 2015). As expected, given their rather wide use in the energy policy domain, the vast majority of MCDM frameworks reviewed were employed in applications in the power sector. The transportation sector also appeared to be popular among multicriteria studies; this can be attributed to the very large number of technological assessment

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

7

Table 2 Overview of climate policy studies in which fuzzy cognitive mapping is integrated with other methodologies. Category

Approach

Study

Communication

Building Block Methodology Delphi

Ceccato (2012) Biloslavo and Dolinšek (2010) Amer et al. (2011) Biloslavo and Grebenc (2012) Kayikci and Stix (2014) Hsueh (2015) Amer et al. (2016) Samarasinghe and Strickert (2013) Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Shiau and Liu (2013) Mourhir et al. (2016) Ortolani et al. (2010) Karavas et al. (2015) Hobbs et al. (2002) Shiau and Liu (2013) Zhao et al. (2014) Huang et al. (2013) Anezakis et al. (2016) van Vliet et al. (2010) Mallampalli et al. (2016) Nikas and Doukas (2016) Mourhir et al. (2016) Meliadou et al. (2012) Singh and Nair (2014) Amer et al. (2011) Amer et al. (2016)

Multiple criteria decision analysis

Computational modelling Statistical analysis

Quantitative modelling

Other

Geomorphic Assessments AHP

TOPSIS Agent-Based Models Multi-Agent Systems Principal Component Analysis

Structural Equation Modelling Climate Models Environmental Models Integrated Assessment Models Driving Forces–Pressures–State–Impact–Responses (DPSIR) Framework Integrated Coastal Zone Management (ICZM) Sustainable Livelihoods Framework (SLF) Technology Roadmapping (TRM)

studies. Only a small number of MCDM applications with climate policy implications regarded the agriculture, building and industry sectors or environmental management. It is noteworthy that one study, focusing on the evaluation of alternative mitigation instruments in different countries across Europe, employed a crosssectoral approach (Konidari & Mavrakis, 2007). The proposed categorisation is economic-sector-oriented and, acknowledging that different approaches for sorting the literature reviewed may be used, significant effort has been put in securing consistency; for example, publications assessing energy efficiency measures are sorted in respective (e.g. buildings or transportation) sector categories. It should also be noted that while the majority of the literature reviewed had climate policy implications, twelve of the studies reviewed were explicitly conducted for the purpose of climate policy support (seven for mitigation, four for adaptation, and one for both). Another significant aspect of multiple criteria analysis, on which the selection of the appropriate MCDM method largely depends, is the type and number of engaged stakeholders it can support. Decision making in complex problems usually involves the selection of one or more solutions among multiple alternatives, evaluated against multiple criteria, by a large number of decision makers or stakeholder groups. The latter’s knowledge of the problem domain may largely vary and their interests do not coincide or may even conflict with each other but should nevertheless be taken into account before reaching a collectively acceptable solution. Stakeholder engagement in climate policy making, in particular, is especially important; as a complex, multi-dimensional and multi-disciplinary process that must result in robust and socially acceptable outcomes, this is a problem domain that would significantly benefit from the involvement of various stakeholder groups in the process. MCDM can help design processes that are inclusive, open, fair and transparent (Phillips & Bana e Costa, 2007), thus significantly contribute to climate action governance. Table 3 also indicates the number and type of stakeholders involved in each of the MCDM studies reviewed, where specified; more than half of these studies did not specify such information, while one study (Chen & Pan, 2015) involved a different number of stakeholders for

the two MCDM stages: weighting of criteria and assessment of alternatives. Regardless of the economic sector of application or explicit reference to climate policy mitigation or adaptation, MCDM can be carried out in different ways. There currently exist a very large number of methodological frameworks, based on different approaches and paradigms, each one with benefits and drawbacks depending on the application scope and purpose. Among the 26 different approaches that were found to have been used in the context of solving problems in this domain, three methods appear to be the most popular: AHP (Saaty, 1990), a structured approach to organising and analysing complex decisions that draws from mathematics and psychology and is based on pairwise comparisons and stakeholder judgment; the PROMETHEE (Brans, Vincke, & Mareschal, 1986) family of methods, another pairwise comparison approach that is based on partial or complete ranking of alternatives; and TOPSIS (Lai, Liu, & Hwang, 1994), a distance-based method that calculates how far each alternative is from the ideal solution. These are followed by the Weighted Sum Method, the outranking ELECTRE (Roy, Présent, & Silhol, 1986) family of methods, the compromise-oriented VIKOR (Opricovic & Tzeng, 2004) method, the fuzzy versions of AHP and TOPSIS, and multi-attribute value theory or MAVT/MAVA (Belton & Stewart, 2002). Multiple-objective programming approaches were also found in the literature, mostly including multi-objective linear programming and goal programming. Table 4 summarises the frameworks found in the review, along with the applications where they were employed and the nature of the evaluation criteria against which alternatives were assessed. As shown in Table 4, many of the studies reviewed employed numerous MCDM frameworks, either in different stages of an integrated approach (superscript “D”) or on a comparative basis (superscript “C”). It should be noted that no decision is taken to classify publications against a specific MCDM method, i.e. to prioritise the methodologies used in each study, when multiple MCDM methods are employed: the studies are hence classified in all relevant MCDM methods in Table 4. For example, Cowan, Daim, and Anderson (2010) use AHP and multi-objective goal programming

8

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Table 3 Overview of climate policy related studies in the MCDM literature, with regard to their application area, sector, explicit policy implications (Superscripts: M=Mitigation, A=Adaptation), and number/type of involved stakeholders (Superscripts: AC=Academics/Researchers, PM=Policymakers, PU=Public, GO=Government officials and ministry representatives, PR=Private sector experts, FI=financial experts, OT=Other or unspecified experts). Application area

Study

Policy evaluation

AlSabbagh et al. (2017)M Batubara et al. (2016) Blechinger and Shah (2011)M Borges and Villavicencio (2004)M Chalabi and Kovats (2014)A Chen and Pan (2015) de Bremond and Engle (2014) A de Bruin et al. (2009)A Georgopoulou et al. (2003)M Javid et al. (2014) Konidari and Mavrakis (2007)M

Sector Agriculture

Project selection

Risk assessment Scenario analysis

Technology assessment

Michailidou et al. (2016)M,A Miller and Belton (2014) A Mourhir et al. (2016) Neves et al. (2008) Oliveira and Antunes (2004) Onu et al. (2017) San Cristóbal (2012) Shiau and Liu (2013) Streimikiene and Baležentis (2013b)M Theodorou et al. (2010) Tsoutsos et al. (2009) Vaillancourt and Waaub (2004) Diakoulaki et al. (2007) Le Téno and Mareschal (1998) Montanari (2004) Perkoulidis et al. (2010) Ramazankhani et al. (2016) Vahabzadeh et al. (2015) Xu et al. (2016) Branco et al. (2012) Baležentis and Streimikiene (2017) Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Jayaraman et al. (2015) Jun et al. (2013) Papadopoulos and Karagiannidis (2008) Almaraz et al. (2013) Antunes et al. (2004) Brand and Missaoui (2014) Büyüközkan and Güleryüz (2017) Büyüközkan and Karabulutb (2017) Chang et al. (2012) Cowan et al. (2010) Cutz et al. (2016) Doukas et al. (2006) Fozer et al. (2017) Ghafghazi et al. (2010) Karakosta et al. (2009) Kaya and Kahraman (2011) Klein and Whalley (2015) Madlener et al. (2009) Maimoun, Madani, and Reinhart (2016) Mohamadabadi et al. (2009) Onar et al. (2015) Paul et al. (2015) Pilavachi et al. (2009) Promentilla et al. (2014) Ren and Lützen (2015) Ribeiro et al. (2013) Rojas-Zerpa and Yusta (2015) Roth et al. (2009) Sadeghi et al. (2012) Sakthivel et al. (2015)

Stakeholders Buildings

Environment

Industry

Power √ √ √

√ √

Transport √

√

√ √

√ √

√

√ √

√

√

√

8PM , 7AC , 8OT 20AC,GO,OT 10OT /25OT

√ √ √

√

√ √ √

√ √

40PM , 400PU

71OT 3 GroupsAC,PM,OT

√ √ √

√

5OT

√ √

√ √

40OT √

√

17GO

√ √ √ √ √

√ √ √ √ √ √ √ √ √ √ √

42OT

3OT 4OT

√

11OT

√ √ √ √ √ √ √ √ √ √ √

√

√

√

17GO , 8PR , 8OT 3OT 3OT 15AC , 3PR 12OT 3 GroupsOT 25OT 3 GroupsPR,PU,OT

√ √ √ √

3OT √ √

√ √ √ √ √

√ √ √ √ √

3OT 50OT 6AC,GO,PR 11AC 16AC,PR,OT 85OT

√ (continued on next page)

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

9

Table 3 (continued) Application area

Study

Sector Agriculture

Other

S¸ engül et al. (2015) Shmelev and van den Bergh (2016) Štreimikiene˙ et al. (2016) Streimikiene et al. (2012) Streimikiene and Baležentis (2013a) Talaei et al. (2014) Ulutas¸ (2005) Volkart et al. (2016) Yap and Nixon (2015) Dace and Blumberga (2016)M Heo et al. (2010) Luthra et al. (2015)

Stakeholders Buildings

Environment

Industry

Power √ √ √ √

Transport

25AC,PR,FI,OT

√ √ √ √ √ √

at different stages and their study is thus classified in both MCDM methods in the table. It is evident that pairwise comparison models like AHP, ANP and Fuzzy AHP are most commonly employed as a first stage of an integrated methodological approach, to define the criteria weights, before employing other MCDM frameworks for the multicriteria analysis of the alternatives. TOPSIS, on the other hand, is frequently used along with other frameworks, like VIKOR, in comparative studies. These findings should be interpreted while acknowledging the limitations of the literature search criteria presented in Section 1, as well as of the proposed classification illustrated in Table 4: for example, some approaches are very similar and yet classified here separately, such as SAW and Weighted Sum Method. The same can be said for SMART and MAVT, or MAUA, which is considered an extension of MAVT that includes probabilities and risk attitudes to form utility functions (Marttunen, Lienert, & Belton, 2017). Depending on the nature (problematic, application area, sector, geographic scope, etc.) of the problem, the criteria of the evaluation are very carefully selected. Modelling a consistent family of evaluation criteria—which are assumed to be non-redundant, exhaustive and cohesive—is a critical process that, according to Roy (1985), takes place at a very early stage, following the determination of the type of problematic (choice, sorting, ranking or description) and the set of different alternatives (Siskos, Grigoroudis, & Matsatsinis, 2005). In MCDM studies with climate policy implications, the selected criteria can be classified into the following dimensions: economic (e.g. debt, economic efficiency, net present value, incentives and subsidies, etc.); energy (e.g. energy intensity, energy efficiency, consumption, contribution to energy independence or security, etc.); environmental and climate-related (e.g. greenhouse gas emissions reduction, mitigation of negative effects to the environment, etc.); regulatory (e.g. accordance or compatibility with the legal framework); social (e.g. societal acceptance, creation of jobs, etc.); technological (e.g. autonomy, reliability, technology transfer capacity, fuel flexibility, etc.); and other criteria (e.g. cooperation capacity, visual impact, etc.). Most of the studies used criteria from multiple dimensions; only a limited number of studies, however, used criteria across all dimensions (Ulutas¸ , 2005; Biloslavo & Grebenc, 2012), while others only focused on the environmental and climate mitigation aspects of their problem ignoring other dimensions. It is noteworthy that both multicriteria decision making and the problems in which it is employed usually feature significant uncertainties. There are multiple ways in which uncertainty is incorporated and dealt with in MCDM applications, with the most prominent one in the literature being sensitivity analysis. In fact, a little less than half of the applications reviewed appeared to perform some kind of sensitivity analysis, as shown in Table 4. Only a limited number of these, however, explicitly referred to the robustness

√ √

50AC,PM,PR

6AC 25OT 3AC

of their results and, if not through the implementation of different MCDM frameworks in the aim of enhancing robustness (Rojas˙ & Zerpa & Yusta, 2015; Streimikiene, Baležentis, Krisciukaitiene, Balezentis, 2012), most of the times this was confused with and performed by means of sensitivity analysis, which, as Durbach and Stewart (2012) point out, usually concerns imprecise judgment issues and is, therefore, a means of addressing “internal” uncertainty. Climate policy, on the other hand, features both risks and uncertainties that are not only limited to the methodological frameworks or to experts’ judgments, but include others that are relevant to the problem domain per se; in other words, they are part of “external” uncertainties and risks that refer to conditions of the problem domain and are outside the control of the decision maker or the modeller. In the MCDM literature, however, these are hardly ever taken directly into account. One popular way of dealing with such uncertainties is the analysis of the range of potential futures by means of different scenarios: Jun, Chung, Kim, and Kim (2013) quantified the flood risk vulnerability in South Korea by considering climate change impacts and comparing various climate change scenarios, Miller and Belton (2014) explored the implications of multiple climate scenarios for informing water resource planning and adaptation policy in Yemen, while Michailidou, Vlachokostas, and Moussiopoulos (2016) explored a climate change mitigation and adaptation strategy in tourist areas through different what-if scenarios. In a different setting and in the aim of dealing with the multi-criteria nature of IAM outputs, Baležentis and Streimikiene (2017) attempted to rank different energy policy scenarios in the EU. Only one study used multicriteria analysis to perform risk assessment: specifically, Branco et al. (2012) employed AHP in order to quantify the exposure level of a selected set of oil companies in the EU to carbon risk. This finding indicates that multicriteria risk assessment is largely underexploited in climate policy studies, despite the fact that quantitatively assessing risks and uncertainties is only feasible in a limited number of relevant aspects of climate policy, which could as a result largely benefit from stakeholder knowledge elicitation and exploitation. Although the vast majority of applications appear not to directly assess these climate change- or policy-related risks and uncertainties, risk management has been indirectly incorporated in certain studies, either as risk-based policies in the set of alternative actions or in the form of risk indices as evaluation criteria (e.g. de Bruin et al., 2009). In other cases, other risk-oriented frameworks were integrated with the overall approach, such as the Political, Economic, Social, Technological, Legal, and Environmental (PESTLE) risk analysis framework or the Benefits–Opportunities–Costs–Risks (BOCR) framework. More insights regarding the different methods, models, frameworks and approaches that have been integrated with MCDM models in the reviewed pieces of literature can be found in Table 5.

10

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Table 4 Overview of the MCDM methodologies employed in the climate policy literature, capturing evaluation criteria (ECO=Economic, ENE=Energy-related, ENV=Environmental and climate-related, REG=Regulatory, SOC=Social, TEC=Technological, OTH = Other), means of engagement (face-to-face interviews/workshops/committees and surveys/questionnaires) and sensitivity analysis. Italic formatting indicates studies with multiple MCDM methods employed, in different (“D” superscript) stages of an integrated approach or for drawing comparisons (“C” superscript) and enhancing robustness. MCDM methodology

AHP

Study

AlSabbagh et al. (2017) Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Blechinger and Shah (2011)D Borges and Villavicencio (2004)D Branco et al. (2012) Büyüközkan and Karabulut (2017)D Cowan et al. (2010)D Javid et al. (2014) Konidari and Mavrakis (2007)D Montanari (2004)D Paul et al. (2015) D,C Pilavachi et al. (2009)D Rojas-Zerpa and Yusta (2015)D

ANP

APIS ARAS DEMATEL ELECTRE

Fuzzy AHP

Fuzzy ANP Fuzzy MCDM Fuzzy PROMETHEE Fuzzy TOPSIS

Fuzzy VIKOR MAUT/MAUA MAVT/MAVA

MOORA MULTIMOORA Multi-objective goal programming

Multi-objective linear programming

Shiau and Liu (2013) Štreimikiene˙ et al. (2016)D Talaei et al. (2014) Theodorou et al. (2010)C Yap and Nixon (2015) Büyüközkan and Güleryüz (2017)D Sakthivel et al. (2015)D Ulutas¸ (2005) Shmelev and van den Bergh (2016) Baležentis and Streimikiene (2017)C Štreimikiene˙ et al. (2016)D Büyüközkan and Güleryüz (2017)D Georgopoulou et al. (2003) Karakosta et al. (2009) Madlener et al. (2009) Michailidou et al. (2016) Neves et al. (2008) Papadopoulos and Karagiannidis (2008) Perkoulidis et al. (2010) Theodorou et al. (2010)C Heo et al. (2010) Kaya and Kahraman (2011)D Luthra et al. (2015) Onar et al. (2015) Ren and Lützen (2015)D Sadeghi et al. (2012)D Promentilla et al. (2014) Chang et al. (2012) Cutz et al. (2016) Chen and Pan (2015) Jun et al. (2013)C Kaya and Kahraman (2011)D Onu et al. (2017) Sadeghi et al. (2012)D S¸ engül et al. (2015) Vahabzadeh et al. (2015) Konidari and Mavrakis (2007)D Chalabi and Kovats (2014) de Bremond and Engle (2014) Miller and Belton (2014) Fozer et al. (2017) Paul et al. (2015) D,C Streimikiene et al. (2012)C Streimikiene and Baležentis (2013b) Cowan et al. (2010)D Jayaraman et al. (2015) San Cristóbal (2012) Antunes et al. (2004) Oliveira and Antunes (2004)

Evaluation criteria ECO √

ENE

ENV √

REG √

SOC √

TEC

OTH √

√ √ √ √ √ √ √ √ √ √

√ √ √

√ √ √ √ √ √ √ √ √ √ √ √ √

√ √ √

√ √ √ √

√ √ √

√ √ √ √ √

√ √

√

√ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √

√

√ √

√

√ √

√

√ √ √ √ √ √ √ √ √ √ √ √

√

√

√

√ √

√ √ √ √ √

√

√ √ √ √ √ √ √ √ √ √ √ √

√ √ √ √

√ √ √ √ √ √ √ √

√

√ √ √

√

√ √ √

Survey, Interviews Survey Survey Survey Workshop

√

Survey Survey Survey, Interviews Committee Survey Workshop

√

√ √ √ √

Survey Interviews Survey

√ √ √ √

Survey Interviews

√ √

√ √

Survey

√ √

√ √ √ √

√ √ √

√

√

Survey, Workshop

√

√ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √

Sensitivity analysis

√

Survey Survey Survey

√

√ √ √ √ √ √ √ √

√ √ √

√ √

Means of engagement

√

√ √

√ √ √ √ √ √ √ √ √ √

√ √ √ √ √ √ √ √ √ √ √

√

√ √

√ √

√ √ √

√ √ √

√

√ √ √

√ √

√

√

√

√

Survey Workshop

Survey Survey Survey Survey

√ √ √ √

√

Survey

√ √ √ √ √ √

√ √ √ Survey

√ √

√

√

√ √ √ √ √

√ √ √

√

√

Survey √ √

(continued on next page)

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

11

Table 4 (continued) MCDM methodology

Study

Evaluation criteria

Ribeiro et al. (2013)D Point allocation method PROMETHEE

SAW SMART TOPSIS

VIKOR

WASPAS Weighted sum method

Xu et al. (2016)D Batubara et al. (2016) Borges and Villavicencio (2004)D Diakoulaki et al. (2007) Doukas et al. (2006) Ghafghazi et al. (2010) Le Téno and Mareschal (1998) Mohamadabadi et al. (2009) Paul et al. (2015)D, C Theodorou et al. (2010)C Tsoutsos et al. (2009) Vaillancourt and Waaub (2004) Xu et al. (2016)D Maimoun et al. (2016)C Blechinger and Shah (2011)D Konidari and Mavrakis (2007)D Almaraz et al. (2013) Baležentis and Streimikiene (2017)C Brand and Missaoui (2014) Büyüközkan and Güleryüz (2017)D Dace and Blumberga (2016) Jun et al. (2013)C Maimoun et al. (2016)C Montanari (2004)D Mourhir et al. (2016) Ramazankhani et al. (2016)C Sakthivel et al. (2015)D,C Streimikiene et al. (2012)C Streimikiene and Baležentis (2013a) Büyüközkan and Karabulut (2017)D Ramazankhani et al. (2016)C Ren and Lützen (2015)D Rojas-Zerpa and Yusta (2015)D Sakthivel et al. (2015)D,C Baležentis and Streimikiene (2017)C de Bruin et al. (2009) C

Jun et al. (2013) Klein and Whalley (2015) Pilavachi et al. (2009)D Ribeiro et al. (2013)D Roth et al. (2009) Volkart et al. (2016)

ECO

ENE

ENV

√

√

√ √ √

√ √

√ √

√ √

√ √ √ √ √ √ √ √ √ √ √ √

√ √

√

√ √

√ √ √ √ √ √ √ √

√ √

√ √

√ √ √ √ √ √

√

√

√

√ √

√

As already discussed, most applications do not solely base their results on one or multiple MCDM frameworks but rather employ more integrated approaches, by combining multicriteria analysis with other methodologies at different stages. Among the many frameworks, models or methodologies that are displayed in Table 5, the most prominent ones are the different communication techniques, like Delphi (e.g. Cowan et al., 2010; Xu, Nayak, Gray, & Ouenniche, 2016) and Input–Output Analysis (e.g. Jayaraman, Colapinto, La Torre, & Malik, 2015). Another finding is that the other two decision support frameworks reviewed in this paper, i.e. FCMs (e.g. Shiau & Liu, 2013) and PA (Almaraz, Azzaro-Pantel, Montastruc, Pibouleau, & Senties, 2013), are integrated with MCDM in different configurations. Most importantly, however, existing approaches integrating MCDA with IAMs, such as WITCH and TIAM (Baležentisa & Streimikiene, 2017), AIM and TIMES (Vaillancourt & Waaub, 2004), and MARKAL (Shmelev & van den Bergh, 2016), prove that climate-economy modelling can benefit from MCDM frameworks, in various integration settings.

SOC

TEC

OTH

√

√

√

√

√

√

√

√ √ √ √ √ √ √ √

√ √ √ √

√ √ √ √ √

√ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ √

REG

√

√ √ √ √ √ √ √

√

√ √

√

√

√

√ √

√ √

√ √ √

√ √ √

√

√

√

√

√

√ √

√ √ √ √

Survey

Survey Survey Survey

Survey

Survey, Interviews Survey √

√ √

√ Survey

√ √

√ √ √

√

√

√

√

√ √ √ √

√

Survey

√ √

√

Workshop

Workshop Interviews

√

√ √ √ √

Workshop

Survey √ √ √

√ √

Sensitivity analysis

√ √ √

√

√

√

Survey, Interviews Survey

√ √ √ √

√

√ √

Means of engagement

√ √

√ √ √

√

Workshop Survey

Survey, Interviews Survey, Interviews

√ √ √

5. Portfolio analysis: transforming climate policy instruments into assets According to Markowitz (1952), selecting the optimal portfolio involves making probabilistic estimates of the future performance of available options, then determining an efficient set of portfolios by means of analysing these estimates, and finally selecting a single portfolio best suited to the investor’s preferences. Although PA corresponds to the second stage of this process (Vörös, 1986), framing a climate policy problem into said context usually involves additional efforts in all three stages. This is probably one of the main reasons why, despite it being around for decades, PA is still relatively underexploited as a climate policy support tool. Nevertheless, 47 pieces of literature were found to tackle problems of the climate policy domain or issues closely linked to climate change mitigation or adaptation policy. Due to the fact that energy investments have occupied research long before climate policy did, however, selecting an

12

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24 Table 5 Overview of climate policy studies in which multiple criteria decision making is integrated with other methodologies. Category

Approach

Study

Communication

Delphi

Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Cowan et al. (2010) Jun et al. (2013) Roth et al. (2009) Xu et al. (2016) Jayaraman et al. (2015) Oliveira and Antunes (2004) San Cristóbal (2012) Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Mourhir et al. (2016) Shiau and Liu (2013) Jun et al. (2013) Shmelev and van den Bergh (2016) Brand and Missaoui (2014) Vaillancourt and Waaub (2004) Baležentis and Streimikiene (2017) Almaraz et al. (2013) Baležentis and Streimikiene (2017) Shmelev and van den Bergh (2016) Ulutas¸ (2005) Yap and Nixon (2015) Madlener et al. (2009) Mourhir et al. (2016) Le Téno and Mareschal (1998) Fozer et al. (2017) Roth et al. (2009) Volkart et al. (2016)

Input–Output Analysis

Semi-quantitative modelling

Fuzzy Cognitive Mapping

Quantitative modelling

Climate Models Energy System Models

Uncertainty assessment

Integrated Assessment Models Portfolio Analysis Monte Carlo Analysis

Other

Benefits–Opportunities–Costs–Risks (BOCR) Framework Data Envelopment Analysis Driving Forces–Pressures–State–Impact–Responses (DPSIR) Framework Life Cycle Analysis

investment portfolio for optimising the power generation mix (e.g. Bar-Lev & Kratz, 1976; or Zhu & Fan, 2010) or the energy mix (e.g. Pugh et al., 2011) of a country or region appears to almost engross the climate policy-related PA literature. Among these studies, many further focused on electricity markets (e.g. Cucchiella, Gastaldi, & Trosini, 2017; Siddiqui, Tanaka, & Chen, 2016; or Zon & Fuss, 2006); while others aimed at promoting renewable energy sources, such as photovoltaics (Albrecht, 2007; Awerbuch, 20 0 0), wind power (Adabi, Mozafari, Ranjbar, & Soleymani, 2016; SantosAlamillos, Thomaidis, Usaola-García, Ruiz-Arias, & Pozo-Vázquez, 2017), and biomass (Lintunen & Uusivuori, 2016). Furthermore, the focus of some of these PA studies was not limited to electricity generation but included the assessment of research and development investments in energy technologies (e.g. Baker & Solak, 2011; Lemoine, Fuss, Szolgayova, Obersteiner, & Kammen, 2012), policy instruments related to emissions trading (Arnesano, Carlucci, & Laforgia, 2012; Flues, Löschel, Lutz, & Schenker, 2014), or measures and technologies in sectors strongly interlinked with the power sector, such as energy efficiency technologies in buildings (Westner & Madlener, 2010; Shakouri, Lee, & Choi, 2015) or transportation fuels (Marrero, Puch, & Ramos-Real, 2015). It is noteworthy that there exist two main portfolio-building approaches in analyses revolving around power generation: capacity-based portfolios and energy-based portfolios; as Jansen, Beurskens, and Van Tilburg (2006) note, while the former may seem closer to the notion of portfolio assets and thus intuitively attractive, the latter are more realistic since installed capacity tends to vary among portfolios having to meet a specific one-period electricity demand. Outside the power sector, PA applications aimed at supporting environmental management and investments in the agriculture and forestry sector. Crowe and Parker (2008) use PA in order to select an optimal set of seed sources towards regenerating forests of white spruce in a climate-resilient manner. Mitter, Heumesser, and Schmid (2015) attempt to quantify climate change impacts on agricultural vulnerability and develop robust crop production portfolios by analysing the effect of agricultural adaptation policies. Mendelsohn and Seo (2007) examine the way farmers choose their

livestock based on different locations in Africa. Lintunen and Uusivuori (2016) seek the optimal mitigation policy portfolio regulating forest carbon flows while also looking at electricity generation from biomass. All environmental PA applications focused on water planning, with adaptation policy implications. Finally, other applications included Lemoine et al. (2012) that look at different research and development options and respective policy instruments towards a dynamic abatement portfolio, Luo and Wu (2016) that examine the global carbon emissions market, and Romejko and Nakano (2017) that focus on alternative fuel vehicles. Regardless of their application area, as many as 21 of these studies explicitly referred to climate change mitigation or adaptation directly, as part of their study scope, by using an emissions-related utility function, or by evaluating climate policy instruments, measures or investments; the remaining studies had indirect implications for climate action, while including environment- and climate-related constraints. Table 6 provides an overview of the studies reviewed, along with their application area and an indication of explicit reference to climate policy As Table 7 suggests, most studies in the relevant literature employ Markowitz’s Modern Portfolio Theory (MPT) or Mean-Variance Analysis (MVA) (Markowitz, 1952) or the Real Options Analysis (ROA) framework (Dixit & Pindyck, 1994). Straying from the traditional route, other known financial, trading and stock valuation approaches in the literature reviewed include the Black–Litterman global portfolio optimisation model (Black & Litterman, 1992), the Clay–Clay Vintage Model (Den Hartog & Tjan, 1980), the Random Walk Theory (Cootner, 1964), and the Capital Asset Pricing Model (CAPM) (Sharpe, 1964; Lintner, 1975). Others primarily include mathematical programming, such as stochastic (e.g. Flues et al., 2014) and quadratic programming (e.g. Lue & Wu, 2007); metaheuristics, such as the Artificial Bee Colony algorithm (Adabi et al., 2016); and statistical approaches, such as the multinomial logistic regression (Mendelsohn & Seo, 2007). Apart from the widespread portfolio optimisation methodologies, certain researchers used different approaches to allocating their assets into optimal portfolios. For instance, Barron, Djimadoumbaye, and Baker (2014) used a

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

13

Table 6 Overview of climate policy related studies in the portfolio analysis literature, with regard to their application area and explicit policy implications (Superscripts: M=Mitigation, A=Adaptation). Study

Application area Upgrading buildings

Adabi et al. (2016) Albrecht (2007) Allan et al. (2011) Arnesano et al. (2012) Awerbuch and Berger (2003) Awerbuch et al. (2006) Awerbuch (20 0 0) Awerbuch (2006) Baker and Solak (2011)M Bar-Lev and Kratz (1976) Barron et al. (2014)M Bhattacharya and Kojima (2012) Buurman and Babovic (2016)A Chalvatzis and Rubel (2015) Crowe and Parker (2008)A Cucchiella et al. (2017) Delarue et al. (2011) M Flues et al. (2014) Fuss et al. (2012)M Hua et al. (2015)A Jansen et al. (2006) Laurikka and Springer (2003)M Lemoine et al. (2012)M Lintunen and Uusivuori (2016)M Liu and Wu (2007) Luo and Wu (2016)M Marinoni et al. (2011)A Marrero et al. (2015) McLohglin and Bazilian (2006) Mitter et al. (2015)A Narita and Klepper (2016) Nazari et al. (2015)A Oda and Akimoto (2011) Pugh et al. (2011)M Romejko and Nakano (2017)M Roques et al. (2008) Santos-Alamillos et al. (2017) Mendelsohn and Seo (2007)A Shahnazari et al. (2017) Shakouri et al. (2015) Siddiqui et al. (2016) Springer (2003)M Torani et al. (2016) van Asseldonk and Langeveld (2007)A Westner and Madlener (2010) White (2007) Zhou et al. (2010)M Zhu and Fan (2010) Zhu and Fan (2011) Ziegler et al. (2012) Zon and Fuss (2006)M Huang and Wu (2008) Muñoz et al. (2009)

Electricity planning √ √ √ √ √ √ √ √ √ √ √ √

Emissions trading

Energy R&D

Clean industry

Agriculture and forestry

Green transport

Water planning

√

√ √ √

√ √ √ √ √ √

√ √

√

√ √

√

√ √

√ √

√ √ √

√ √

√ √

√

√ √ √ √

√ √ √

√

√ √

√ √ √ √

√ √ √ √ √

√

√ √

√

√ √ √ √ √ √ √ √ √

budget-constrained model, for minimising the expected social cost by choosing the optimal R&D portfolio, and compared it with a two-stage overall optimisation model that not only allocates the budget optimally but also features the capacity to determine the optimal size of the R&D budget. A significant dimension of PA, which probably hinders its further diffusion in this domain, lies in the determination of the financial assets to invest in. As Liu (1999) notes, assets are random variables with a probability distribution over their possible returns, a linear combination of which comprises a portfolio. Given the difficulty in characterising any policy instrument as an asset, most applications in this field have expectedly focused on assessing investments instead of other policy-related measures; this explains why power generation problems compose the vast major-

√

ity of these applications. Nevertheless, some studies also focused on non-investment assets, such as financial incentives (e.g. government contracts in Roques, Newbery, & Nuttall, 2008; and subsidies in Bhattacharya & Kojima, 2012), carbon emission taxes (Awerbuch, 20 0 0; Zhu & Fan, 2010; Lintunen & Uusivuori, 2016), renewable portfolio standards and cap-and-trade schemes (Siddiqui et al., 2016), or seed sources and crop yields in agriculture studies (e.g. Crowe & Parker, 2008). The most central aspect of PA is the problem modelling process. This includes the identification of the objective or utility function, the optimisation of which must be achieved in order to secure the range (frontier) of efficient portfolios, and the restrictions or constraints that this function is subject to. The vast majority of the studies reviewed employ methodological approaches based on a

14

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Table 7 Overview of PA approaches, assets (F = Financial Incentives, I = Investments, P = Policies, T = Taxes, O = Other), objective functions, constraints (ECO = Economic, ENV = Environmental or Climate, GEO = Geographical, POL = Political, REG = Regulatory, TECH = Technological, TIME = Time-related) and uncertainty assessment methods in climate policy studies. Approaches

Study

Assets

Black–Litterman Model Arnesano et al. (2012) Capital Asset Pricing Model Albrecht (2007) Marrero et al. (2015)

P I I; P

Clay–Clay-Vintage Model

Zon and Fuss (2006)

I

Modern Portfolio Theory

Allan et al. (2011)

I

Multi-Criteria Diversity Analysis Random Walk Theory Real Options Analysis

Utility function(s)

Cost minimisation Risk minimisation Cost and Risk minimisation Cost and Risk minimisation Cost and Risk minimisation Cost minimisation

Awerbuch and Berger (2003) Awerbuch et al. (2006)

P

Awerbuch (2006) Bar-Lev and Kratz (1976)

P I

Bhattacharya and Kojima (2012) Chalvatzis and Rubel (2015) Crowe and Parker (2008) Cucchiella et al. (2017)

F; I

Delarue et al. (2011) Hua et al. (2015) Jansen et al. (2006)

I I I

Liu and Wu (2007)

I

Luo and Wu (2016) Marinoni et al. (2011) Marrero et al. (2015)

I; P I I; P

McLohglin and Bazilian (2006) Mitter et al. (2015) Nazari et al. (2015) Roques et al. (2008) Santos-Alamillos et al. (2017) Shakouri et al. (2015)

P

Risk minimisation Risk minimisation Profit maximisation; Risk minimisation Cost minimisation Risk minimisation Cost and Risk minimisation Profit maximisation; Risk minimisation Risk minimisation Profit maximisation Cost and Risk minimisation Cost minimisation

P I; P F; I I

Profit maximisation Cost minimisation Profit maximisation Risk minimisation

I

Capacity and profit maximisation; Risk minimisation Produce maximisation

I

I O I

van Asseldonk and Langeveld (2007) Westner and Madlener (2010) White (2007)

O

Zhu and Fan (2010) Huang and Wu (2008)

I; T I

Awerbuch et al. (2006)

P

Ziegler et al. (2012)

I

I; P I

Buurman and Babovic (2016) Fuss et al. (2012)

I; P

Narita and Klepper (2016) Nazari et al. (2015) Oda and Akimoto (2011) Shahnazari et al. (2017)

I I; P I

Torani et al. (2016) Zhou et al. (2010) Zhu and Fan (2011)

I

F; I; T I I

Capacity and diversity maximisation; risk minimisation Cost minimisation Cost and Risk minimisation; Risk minimisation

Constraints ECO √

Cost and Risk minimisation Profit maximisation Cost minimisation Profit maximisation Profit maximisation; Risk minimisation Cost minimisation Profit maximisation Profit maximisation

Uncertainty GEO

POL

REG

TECH TIME

√ √

√

√ √

√

Scenarios

√

Scenarios

√

√

√

√

√ √

√ √

√

√ √ √ √

√ √ √

√ √

√

√ √

√

√ Scenarios

√

Scenarios; CVaR Scenarios

√ √ √

√ √ √

√

Scenarios CVaR Monte Carlo Scenarios

√

√

√

Scenarios √

Profit maximisation; Risk minimisation Cost and Risk minimisation Cost minimisation Cost and Risk minimisation Cost minimisation Profit maximisation; Risk minimisation Cost minimisation

ENV

√

Monte Carlo

√

√ √

√

√

√

√

√

Scenarios

√

Cholesky decomposition

√ √ √

√

√ √

√ √

√ √ √

√ √ √

√ √ √

√ √ √

√ √

√ √

Scenarios; Monte Carlo; CVaR CVaR Scenarios; Monte Carlo; CVaR Scenarios Scenarios; Monte Carlo Scenarios; Monte Carlo (continued on next page)

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

15

Table 7 (continued) Approaches

Study

Assets

Utility function(s)

Constraints ECO

Other

Adabi et al. (2016) Awerbuch (20 0 0) Baker and Solak (2011) Bar-Lev and Kratz (1976)

I I; T I I

Barron et al. (2014)

I

Flues et al. (2014) Laurikka and Springer (2003) Lemoine et al. (2012) Lintunen and Uusivuori (2016) Liu and Wu (2007) Pugh et al. (2011) Romejko and Nakano (2017) Mendelsohn and Seo (2007) Siddiqui et al. (2016)

I I I P

ENV

POL

REG

TECH TIME

√

√

I I

Risk minimisation Profit maximisation Cost minimisation Cost and Risk minimisation Social cost minimisation; Budget optimisation Cost minimisation Emissions minimisation

P T

Cost miminisation Cost minimisation

√

√ √

I

Profit maximisation; Risk minimisation Cost minimisation Cost minimisation Profit maximisation Social cost minimisation

√

√

cost or risk minimisation utility function, although a large number of them employ more than one utility functions at different stages. Fewer applications aimed at profit or return maximisation instead of cost minimisation—some researchers suggest that return maximisation is the same as the reciprocal of cost minimisation (e.g. Awerbuch & Berger, 2003), while others define these two as not essentially the same if properly defined (Jansen et al., 2006). Regarding constraints, although as many as seven different categories were identified (financial; environmental or climate; geographical; political; regulatory or policy; technological; and time-related), they appeared to quite vary. The most common ones included budget or market constraints, upper boundaries in technological shares in energy mixes, emission constraints, limitations drawing from EU directives and national policies, time horizons, and location distance combinations. Last but not least, a major feature of PA—and probably the one with the highest potential for climate change and policy design—is linked to its inherent capacity to deal with risks and uncertainties. Aside from the standard integration of risk in the MVA approach, many researchers attempted to minimise risk as part of their objective functions, or by means of the conditional valueat-risk (CVaR) tool, i.e. the expected average shortfall, (e.g. Nazari, Maybee, Whale, & McHugh, 2015). The most prevalent uncertainty assessment method in portfolio analysis is the use of different scenarios driving the modelling process. For instance, Fuss, Szolgayová, Khabarov, and Obersteiner (2012) made use of the shared socio-economic pathways (O’Neill et al., 2014), while Marinoni, Adkins, and Hajkowicz (2011) and Hua et al. (2015) assumed different climate change scenarios for their adaptation policy portfolio analyses. A generally widespread approach to uncertainty assessment is Monte Carlo analysis, which—contrary to scenario analyses— samples values from probability distributions for multiple variables, in order to produce numerous combinations, against all of which the model is stress-tested towards reaching the most robust set of portfolios. Table 7 summarises the methodological approaches used in the climate policy literature, along with the nature of assets modelled, the utility functions, the type of constraints to which these functions were subject, as well as the method of uncertainty assessment. Aside from any methodological implications regarding PA, only a small number of studies integrated it with other tools. In particular, analyses in the agricultural and forestry applications employed

Uncertainty GEO

√ √ √

√

√

√

Monte Carlo

√

√

√ √ Scenarios

√ √

√ √

√ √

Scenarios; Monte Carlo Scenarios Scenarios

different modelling frameworks, including a species-range impact model (Crowe & Parker, 2008), a crop rotation model (Mitter et al., 2015) and the WOFOST crop simulation model (van Asseldonk & Langeveld, 2007). Finally, Luo and Wu (2016) used the orthogonal GARCH model to examine the time-varying correlations in EU carbon allowance, crude oil and stock markets in Europe, China and the US; while Marinoni et al. (2011) used MCDM and MPT to determine a portfolio of intervention sites that returns maximum attainable aggregated benefit for a fixed budget. Most importantly for the purposes of this review, two studies integrated IAMs with portfolio approaches, namely Pugh et al. (2011) that used technological scenarios from the GCAM model and Baker and Solak (2011) that used outputs from the DICE2007 and MiniCAM models. 6. An example of an integrative scientific approach FCMs, MCDM and PA reportedly appear to provide policy makers with less complex, stakeholder-driven tools for assessing hardto-model dimensions of the problem domain. As the reviewed literature indicates, all of these frameworks appear to have a good methodological integration record and have so far supported climate policy making, either directly or indirectly, across a number of application areas and economic sectors. This cultivates the ideal framing for shaping a new, integrative approach, in which energy and climate-economy modelling and decision support frameworks bring different bits and pieces of vital importance to the table, thus putting together a far-reaching jigsaw of aspects that must be taken into account in climate policy making. The idea of employing different modelling approaches, in the face of uncertainty, in order to enhance climate policy-oriented modelling processes is not new (e.g. see Lange & Treich, 2008). Based on the findings of Sections 3–5, multiple settings of approaches have been identified in applications across the climate policy literature, integrating the methodologies reviewed with climate-economy and/or energy modelling, or with one another. These applications are depicted in Table 8. For the purposes of this study, an illustrative example of integrating these three decision support frameworks with a modelling ensemble has been elaborated (Fig. 1), as a real-world climate policy problem. In order to achieve the near-term national energy efficiency targets in Greece, as established in EU’s Directive 2012/27/EC, and in line with the country’s long-term efforts for climate change

16

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24 Table 8 Approaches integrating IAMs, FCMs, MCDM and PA in the reviewed literature. Study

IAM(s)

Almaraz et al. (2013) Baker and Solak (2011) Baležentis and Streimikiene (2017) Biloslavo and Dolinšek (2010) Biloslavo and Grebenc (2012) Marinoni et al. (2011) Mourhir et al. (2016) Pugh et al. (2011) Shiau and Liu (2013) Shmelev and van den Bergh (2016) Vailliancourt and Waaub (2004)

FCM

DICE2007, MiniCAM WITCH, TIAM

√ √ √

GCAM

√

MARKAL AIM, TIMES

MCDM

PA

TOPSIS

Linear programming StochASTIC OPTIMisation

ARAS, TOPSIS, WAPSAS AHP AHP Compromise programming TOPSIS

Modern portfolio theory Rank-ordered ROI

AHP APIS PROMETHEE

MCDM TOPSIS

Risk indices

Modelling ensemble TIMES/WASP-IV

Costs and energy savings

PA AUGMECON-2

Near-optimal policy portfolios

FCM

Ranking of policy mixes

Policy recommendations

Fig. 1. Integrating decision support frameworks and quantitative system models in a climate policy problem.

mitigation, a modelling ensemble comprising TIMES (Loulou, Remme, Kanudia, Lehtila, & Goldstein, 2005) and WASP-IV (International Emissions Trading Association, 2006) was used to analyse and predict national consumption per energy product, sector and use. By decodifying the energy savings and investment cost results of the TIMES/WASP-IV modelling runs in the form of a selected mix of technologies, as presented in Ministry of the Environment and Energy (2017), fifteen policy measures for the building and transportation sectors were identified: P1. The “Save Energy at Home II” financing mechanism; P2. Energy upgrade of buildings of the broader public sector; P3. Energy efficiency and demonstration projects in Small- and Medium-sized Enterprises (SMEs); P4. Implementation of the “ISO 50001” energy management system in the public sector; P5. Energy upgrade of commercial buildings through ESCOs; P6. Deployment of smart metering systems; P7. Implementation of the EPPERRA (Environment and Sustainable Development) programme; P8. Offsetting fines on illegal buildings with energy-upgrading interventions; P9. Energy managers in the broader public sector; P10. District heating network design and development; P11. Replacement of old public and private light trucks; P12. Replacement of old private passenger vehicles; P13. Improvements in road lighting; P14. Development and improvements of pump stations; and P15. Proactive promotion of Energy Performance Certificates (EPCs). Previous legislative failures and resulting divergence from the national energy efficiency targets, however, showed that various uncertainties and risks hindering the successful implementation of these or similar instruments had been overlooked in the past (Nikas, Gkonis et al., 2019). Aiming to capture the knowledge embedded in policy makers, regarding the impact of such risks on the energy efficiency policy framework, a risk index was co-defined for each of the fifteen instruments. Specifically, seven policy makers from the Ministry were engaged in interviews and asked to evaluate each policy instrument, in a structured questionnaire, against eight implementation risks: R1. Difficulties in aligning local authorities with national obligations; R2. Political instability; R3. Bureaucracy; R4. Demanding regulatory framework; R5. Inadequate banking sector; R6. Social opposition; R7. Unqualified personnel; and R8. Poor

market conditions. By employing a TOPSIS-based MCDM approach, a risk index was calculated for each policy measure, as depicted in Fig. 2. In particular, a group decision making variant of Fuzzy TOPSIS (Chen, 20 0 0) was used, by applying Fuzzy TOPSIS once for each of the seven policy makers, and once more for the global preference model that is based on the aggregation of the results of the seven individual preference models, as suggested by Krohling and Campanharo (2011). The final results of the group Fuzzy TOPSIS, expressing the group relative closeness of each policy instrument to the ideal solution, were used as a risk index for the 15 policy measures. For example, as suggested by Fig. 2, offsetting fines on illegal buildings with energy upgrades (P8) is considered by the engaged policy makers to be the riskiest among the considered policy instruments and, based on the results, is assigned a risk index of 0.812; while replacing old public and private light trucks (P11) is considered to be the safest one, in terms of the eight implementation risks, and is assigned a risk index of 0.319. The fifteen policy measures, along with (a) the TIMES/WASP-IV outputs for their achieved energy savings and associated costs and (b) their risk indices as resulted in the implementation of TOPSIS, were then used in a portfolio analysis, with two objective functions: one for risk minimisation and another one for energy savings maximisation, for a given budget available by the Greek Ministry of the Environment and Energy. In order to carry out this multi-objective optimisation, the augmented ε -constraint method (Mavrotas & Florios, 2013) was used, allowing for the assessment of the robustness of both policy portfolios and individual policy instruments (Mavrotas, Figueira, & Siskos, 2015) and the prospects of enhanced visualisation and easy communication of the results to policy makers (Siskos & Tsotsolas, 2015). The portfolio analysis component led to the Pareto front of near-optimal portfolios depicted in Fig. 3 (vertical axis: normalised risk; horizontal axis: energy savings), none of which included investments in policies P2 (energy upgrade of buildings of the broader public sector), P7 (implementation of the Environment and Sustainable Development programme) and P10 (district heating network design and development).

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

17

Fig. 2. Risk indices for the fifteen policy instruments, as resulted from the TOPSIS framework (Forouli et al., 2019).

Pareto front of near-optimal policy portfolios 0.8

Normalised Risk

0.6

0.4

0.2

14.90659 2.824769 612.8786

0 200

400

600

800 Energy savings (kTOE)

1000

1200

Fig. 3. Pareto front of near-optimal policy portfolios, based on the PA approach (Forouli et al., 2019). The four highlighted points correspond to the four selected policy portfolios (green, purple, blue and grey for policy mixes PX1, PX2, PX3 and PX4). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The detailed results of the MCDM and PA components are presented and discussed in Forouli, Nikas, and Doukas (2019). Although the PA results provide useful insights into both the expected energy savings and the associated risk of each policy portfolio, the resulting Pareto front essentially indicates a diversity of near-optimal solutions among which the policy makers may choose, towards maximising energy savings in the safest setting possible. The selection of the specific policy portfolio thus remains up to policy makers, based on their understanding, political perspective and risk aversion levels, a choice that can prove challenging. The FCM methodology was thus selected, in order to facilitate policy makers’ selection of the optimal portfolio. In a dedicated workshop, the 7 policy makers were asked to design the FCM resulting to the model illustrated in Fig. 4. The resulting FCM model was simulated for different policy portfolios comprising the most robust policy measures (M1–M15), against different scenarios designed for different extent of the eight implementation risks (R1–R8). Indicatively, four policy portfolios comprising eight measures were selected along the Pareto front, so as to include different levels of risks and resulting efficiency; these are illustrated in Fig. 3 as highlighted points, with different colours (green, purple, blue and grey points, corresponding to Policy Mixes PX1, PX2, PX3 and PX4, respectively). The portfolios were quantified in [0, 1], according to their financial con-

Table 9 FCM rankings of the four policy mixes for each of the assumed socioeconomic scenarios. FCM scenarios

Rankings

No risks assumed Scenario 1: Sustainability Scenario 2: Middle of the road Scenario 3: Regional rivalry Scenario 4: Inequality Scenario 5: Fossil-fuelled development

PX4 PX4 PX2 PX2 PX2 PX2

> > > > > >

PX2 PX3 PX3 PX3 PX3 PX3

> > > > > >

PX3 PX2 PX4 PX4 PX4 PX4

> > > > > >

PX1 PX1 PX1 PX1 PX1 PX1

tribution in the portfolio (Fig. 5); the eight implementation risks were also quantified, in five different scenarios inspired by the socioeconomic descriptions discussed in O’Neill et al. (2017), as suggested in Nikas, Ntanos, and Doukas (2019) and Nikas, Stavrakas et al. (2019). Results showed that, from the stakeholders’ perspective, the most robust (across the five scenarios) policy portfolio was policy mix PX2 (Table 9), i.e. the mix corresponding to the second among the selected portfolios. This portfolio comprises large investments in diffusing smart metering systems in Greek households, intermediate investments in financial incentives in both the residential and the private services sector, as well as a small budget allocation to the appointment of energy managers in the broader public

18

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Fig. 4. The FCM, as designed by the engaged policy makers. Arrow thickness and color indicate weight and sign (red = negative, green = positive). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article .)

Fig. 5. Synthesis of the four selected near-optimal policy mixes (portfolios).

sector. The results, furthermore, showed that risk perception has significant impacts on the ranking of the policy mixes: the riskier the system is perceived to be, in terms of challenges, the worse the diversification of the policy framework performs. They also indicated that the seven Greek policy makers appeared to be riskaverse, but also to prefer robust, relatively diverse policy portfolios comprising a small number of policy measures over portfolios with single-policy investments. On the other hand, the most diverse policy portfolios appear to perform better, from the policy makers’ perspective, in socioeconomically optimistic scenarios; this can be attributed to the large number of risks, to which these portfolios are normally vulnerable but the effect of which is negligible in scenarios with low expected challenges. This illustrative example shows how a climate policy problem can be solved by means of a new scientific paradigm that is based on coupling quantitative models with bottom-up decision support frameworks. Drawing from examples in the recent literature (Table

8), it illustrates how modelling outcomes on specific policies or technologies can be assessed in a PA approach that evaluates them against multiple objectives, one of which can be policy-related risks determined in an MCDM approach. FCM can be further used to help policy makers select among the diverse set of resulting near-optimal portfolios, on the basis of reference socioeconomic scenarios that are frequently used by IAM modellers in climateeconomy and policy analysis studies. Such a paradigm can enhance the legitimacy of the scientific processes in support of climate policymaking, by introducing a dynamic stakeholder engagement framework, thereby improving the transparency of the employed models, methods and tools: policy makers and potentially other stakeholder groups can be involved in multiple stages, from the formulation of the policy and research questions driving the quantitative modelling exercises to the assessment of various dimensions of the examined policies, and eventually to the final selection among the resulting near-

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

optimal policy prescriptions. It also allows for the mobilisation of tacit knowledge embedded in experts, towards bridging knowledge gaps and improving the robustness of modelling outcomes against different types of uncertainty, with which climate change and action are inevitably intertwined. Finally, the proposed approach promotes interdisciplinary perspectives and deep uncertainty assessment towards delivering more effective strategies for the realisation of the required transformations (Scrieciu, Belton, Chalabi, Mechler, & Puig, 2014), and shows that operational research can be a critical component of the science underpinning the design of effective, robust, socially acceptable, financially viable and technically feasible climate action. 7. Conclusions Achieving low carbon transitions is a rather complex, multidisciplinary process that does not only involve developing longterm concentration and socio-economic pathways but also requires assessing the policy instruments, strategies and mixes that can promote these pathways in a robust, socially acceptable and viable manner. Therefore, climate policy making must involve equally diverse scientific fields; in this respect, most climate policy studies utilise IAMs, which however feature major weaknesses. There exist decision support frameworks that can address these weaknesses, which can be summarised in (a) small participation of stakeholders, and (b) limited capacity for assessing risks and uncertainties, and thereby support the policy making process by complementing IAMs. This paper reviews three such frameworks, namely FCMs, MCDM and PA, and investigates their existing relationship with actual problems of the climate policy domain. The authors argue that climate policy making can significantly benefit from integrated approaches, in which different methodologies based on the reviewed frameworks complement each other towards reaching robust results of wider acceptance and trust. In particular, given their stakeholder-driven process, FCMs constitute an appropriate framework for comparing and evaluating alternative options against each other in terms of what appears to be most effective or feasible from the experts’ point of view. In this direction and considering that most integrated assessment studies can, based on specific assumptions, provide general insights into policy pathways, FCMs can be used as a complementary framework for evaluating different policy strategies for promoting these pathways, as the recent literature suggests (e.g. Nikas & Doukas, 2016; Mallampalli et al., 2016). Alternatively, as a powerful stakeholder knowledge elicitation framework, they can be used in climate policy studies in stages preceding the use of IAMs (as suggested by van Vliet et al., 2010), for capturing the context of a study while bridging modellers with stakeholders. MCDM, on the other hand, can effectively be used for focusing on specific aspects of the climate policy problem domain, which require comparing multiple alternative actions (such as policy strategies) against multiple evaluation criteria, often based on multiple stakeholders’ opinions. MCDM can also be used for comparing and evaluating IAM-based resulting scenarios, as discussed in Baležentis and Streimikiene (2017). More importantly, climate policy making can significantly benefit from multiple criteria risk assessment approaches (e.g. Branco et al., 2012), and complement IAM-driven studies by addressing both of their major weaknesses, i.e. limited participation of stakeholders and gaps in policy-related risk and uncertainty evaluation, through stakeholder-driven risk assessment studies. Finally, PA can be utilised as a meta-analysis approach, following IAM-driven studies, be delving into economic relations towards determining the optimal combination of assets, such as investments, technologies, carbon taxes, subsidies and other financially quantifiable policy instruments. Its inherent capacity to emphasise

19

robustness, by means of scenario building and the exploitation of probabilistic distributions for a number of parameters, can make it a valuable tool for stress-testing modelling results and focusing on uncertainty-related aspects of climate policy, towards putting together portfolios of policy instruments that are less vulnerable to the most possible future developments (e.g. Pugh et al., 2011). Towards validating this suggestion, the authors presented an integrative scientific approach, though a real-world climate policy problem, in which the methodologies reviewed (MCDM, PA, and FCM) are used to transform climate-economy modelling results into crisp policy recommendations. As these decision support frameworks are integrated, policy makers are all the more involved in and better supervise the process, and results are refined and stress-tested against exogenous risks, leading to better shaped and robust policy recommendations. Perspectives for further research should include customised approaches to appreciating the critical differences of each IAM category—as explored in Section 2—and, to a finer detail, valuing the unique characteristics of each climate-economy model, towards developing tailor-made integrated frameworks with effective FCM, MCDM and PA methodologies to fit the needs of each IAM and case study focus. Acknowledgement The most important part of this research is based on the H2020 European Commission Project “Transitions pathways and risk analysis for climate change mitigation and adaptation strategies— TRANSrisk” under grant agreement No. 642260. The sole responsibility for the content of this paper lies with the authors. The paper does not necessarily reflect the opinion of the European Commission. References Adabi, F., Mozafari, B., Ranjbar, A. M., & Soleymani, S. (2016). Applying portfolio theory-based modified ABC to electricity generation mix. International Journal of Electrical Power & Energy Systems, 80, 356–362. Agrawala, S., Bosello, F., Carraro, C., de Bruin, K., De Cian, E., Dellink, R., et al. (2010). Plan or react? Analysis of adaptation costs and benefits using integrated assessment models. OECD Environment Working Papers, (23), 0_1. Albrecht, J. (2007). The future role of photovoltaics: A learning curve versus portfolio perspective. Energy Policy, 35(4), 2296–2304. Allan, G., Eromenko, I., McGregor, P., & Swales, K. (2011). The regional electricity generation mix in Scotland: A portfolio selection approach incorporating marine technologies. Energy Policy, 39(1), 6–22. Almaraz, S. D. L., Azzaro-Pantel, C., Montastruc, L., Pibouleau, L., & Senties, O. B. (2013). Assessment of mono and multi-objective optimization to design a hydrogen supply chain. International Journal of Hydrogen Energy, 38(33), 14121–14145. AlSabbagh, M., Siu, Y. L., Guehnemann, A., & Barrett, J. (2017). Integrated approach to the assessment of CO2 e-mitigation measures for the road passenger transport sector in Bahrain. Renewable and Sustainable Energy Reviews, 71, 203–215. Amer, M., Daim, T. U., & Jetter, A. (2016). Technology roadmap through fuzzy cognitive map-based scenarios: The case of wind energy sector of a developing country. Technology Analysis & Strategic Management, 28(2), 131–155. Amer, M., Jetter, A., & Daim, T. (2011). Development of fuzzy cognitive map (FCM)-based scenarios for wind energy. International Journal of Energy Sector Management, 5(4), 564–584. Anezakis, V. D., Dermetzis, K., Iliadis, L., & Spartalis, S. (2016). Fuzzy cognitive maps for long-term prognosis of the evolution of atmospheric pollution, based on climate change scenarios: The case of athens. In Proceedings of the international conference on computational collective intelligence (pp. 175–186). Springer International Publishing. Antunes, C. H., Martins, A. G., & Brito, I. S. (2004). A multiple objective mixed integer linear programming model for power generation expansion planning. Energy, 29(4), 613–627. Arnesano, M., Carlucci, A. P., & Laforgia, D. (2012). Extension of portfolio theory application to energy planning problem – The Italian case. Energy, 39(1), 112–124. Ashnani, M. H. M., Miremadi, T., Johari, A., & Danekar, A. (2015). Environmental impact of alternative fuels and vehicle technologies: A life cycle assessment perspective. Procedia Environmental Sciences, 30, 205–210. Auvinen, H., Ruutu, S., Tuominen, A., Ahlqvist, T., & Oksanen, J. (2015). Process supporting strategic decision-making in systemic transitions. Technological Forecasting and Social Change, 94, 97–114.

20

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Awerbuch, S. (20 0 0). Investing in photovoltaics: Risk, accounting and the value of new technology. Energy Policy, 28(14), 1023–1035. Awerbuch, S. (2006). Portfolio-based electricity generation planning: Policy implications for renewables and energy security. Mitigation and Adaptation Strategies for Global Change, 11(3), 693–710. Awerbuch, S., & Berger, M. (2003). Applying portfolio theory to EU electricity planning and policy making. EET IAEA/EET Working Paper No. 03. Awerbuch, S., Stirling, A., Jansen, J. C., & Beurskens, L. W. (2006). Full-spectrum portfolio and diversity analysis of energy technologies. In Managing Enterprise Risk (pp. 202–222). Elsevier Science Ltd. Baker, E., & Solak, S. (2011). Climate change and optimal energy technology R&D policy. European Journal of Operational Research, 213(2), 442–454. Baležentis, T., & Streimikiene, D. (2017). Multi-criteria ranking of energy generation scenarios with Monte Carlo simulation. Applied Energy, 185, 862–871. Barker, T., & Zagame, P. (1995). E3ME: An energy-environment-economy model for Europe. Brüssel: European Commission. Barker, T., Pan, H., Köhler, J., Warren, R., & Winne, S. (2006). Decarbonizing the global economy with induced technological change: Scenarios to 2100 using E3MG. The Energy Journal, 27, 241–258. Bar-Lev, D., & Katz, S. (1976). A portfolio approach to fossil fuel procurement in the electric utility industry. The Journal of Finance, 31(3), 933–947. Barron, R., Djimadoumbaye, N., & Baker, E. (2014). How grid integration costs impact the optimal R&D portfolio into electricity supply technologies in the face of climate change. Sustainable Energy Technologies and Assessments, 7, 22–29. Batubara, M., Purwanto, W. W., & Fauzi, A. (2016). Proposing a decision-making process for the development of sustainable oil and gas resources using the petroleum fund: A case study of the East Natuna gas field. Resources Policy, 49, 372–384. Behzadian, M., Kazemzadeh, R. B., Albadvi, A., & Aghdasi, M. (2010). PROMETHEE: A comprehensive literature review on methodologies and applications. European Journal of Operational Research, 200(1), 198–215. Belton, V., & Stewart, T. (2002). Multiple criteria decision analysis: An integrated approach. Springer Science & Business Media. Benestad, R. E., Nuccitelli, D., Lewandowsky, S., Hayhoe, K., Hygen, H. O., van Dorland, R., et al. (2016). Learning from mistakes in climate research. Theoretical and Applied Climatology, 126(3–4), 699–703. Bhattacharya, A., & Kojima, S. (2012). Power sector investment risk and renewable energy: A Japanese case study using portfolio risk optimization method. Energy Policy, 40, 69–80. Biloslavo, R., & Dolinšek, S. (2010). Scenario planning for climate strategies development by integrating group Delphi, AHP and dynamic fuzzy cognitive maps. Foresight, 12(2), 38–48. Biloslavo, R., & Grebenc, A. (2012). Integrating group Delphi, analytic hierarchy process and dynamic fuzzy cognitive maps for a climate warning scenario. Kybernetes, 41(3/4), 414–428. Black, F., & Litterman, R. (1992). Global portfolio optimization. Financial Analysts Journal, 48(5), 28–43. Blechinger, P. F. H., & Shah, K. U. (2011). A multi-criteria evaluation of policy instruments for climate change mitigation in the power generation sector of Trinidad and Tobago. Energy Policy, 39(10), 6331–6343. Bollen, J. C., & Gielen, A. M. (1999). Economic impacts of multilateral emission reduction policies: Simulations with WorldScan. In Proceedings of the international environmental agreements on climate change (pp. 155–167). Springer. Borges, P. C., & Villavicencio, A. (2004). Avoiding academic and decorative planning in GHG emissions abatement studies with MCDA: The Peruvian case. European Journal of Operational Research, 152(3), 641–654. Bosello, F., De Cian, E., Eboli, F., & Parrado, R. (2009). Macro-economic assessment of climate change impacts: A regional and sectoral perspective. Impacts of climate change and biodiversity effects. European Investment Bank, University Research Sponsorship Programme Final report of the CLIBIO project. Branco, D. A. C., Rathmann, R., Borba, B. S. M., de Lucena, A. F. P., Szklo, A., & Schaeffer, R. (2012). A multicriteria approach for measuring the carbon-risk of oil companies. Energy Strategy Reviews, 1(2), 122–129. Brand, B., & Missaoui, R. (2014). Multi-criteria analysis of electricity generation mix scenarios in Tunisia. Renewable and Sustainable Energy Reviews, 39, 251– 261. Brans, J. P., Vincke, P., & Mareschal, B. (1986). How to select and how to rank projects: The PROMETHEE method. European Journal of Operational Research, 24(2), 228–238. Brown, S. M. (1992). Cognitive mapping and repertory grids for qualitative survey research: Some comparative observations. Journal of Management Studies, 29(3), 287–307. Brundtland Commission. (1987). Our common future: The world commission on environment and development. United Nations: Oxford University Press ISBN: 019282080X. Buonanno, P., Carraro, C., & Galeotti, M. (2003). Endogenous induced technical change and the costs of Kyoto. Resource and Energy Economics, 25(1), 11– 34. Buurman, J., & Babovic, V. (2016). Adaptation pathways and real options analysis: An approach to deep uncertainty in climate change adaptation policies. Policy and Society, 35(2), 137–150. Büyüközkan, G., & Güleryüz, S. (2017). Evaluation of renewable energy resources in Turkey using an integrated MCDM approach with linguistic interval fuzzy preference relations. Energy, 123, 149–163. Büyüközkan, G., & Karabulut, Y. (2017). Energy project performance evaluation with sustainability perspective. Energy, 119, 549–560.

Carraro, C., Galeotti, M., & Gallo, M. (1996). Environmental taxation and unemployment: Some evidence on the ‘double dividend hypothesis’ in Europe. Journal of Public Economics, 62(1), 141–181. Ceccato, L. (2012). Three essays on participatory processes and integrated water resource management in developing countries. Università Ca’ Foscari Venezia, Venice. Çelik, F. D., Ozesmi, U., & Akdogan, A. (2005). Participatory ecosystem management planning at Tuzla Lake (Turkey) using fuzzy cognitive mapping. arXiv preprint q-bio/0510015. Chalabi, Z., & Kovats, S. (2014). Tools for developing adaptation policy to protect human health. Mitigation and Adaptation Strategies for Global Change, 19(3), 309–330. Chalvatzis, K. J., & Rubel, K. (2015). Electricity portfolio innovation for energy security: The case of carbon constrained China. Technological Forecasting and Social Change, 100, 267–276. Chang, P. L., Hsu, C. W., & Lin, C. Y. (2012). Assessment of hydrogen fuel cell applications using fuzzy multiple-criteria decision making method. Applied Energy, 100, 93–99. Chen, C. T. (20 0 0). Extensions of the TOPSIS for group decision-making under fuzzy environment. Fuzzy Sets and Systems, 114(1), 1–9. Chen, L., & Pan, W. (2015). A BIM-integrated fuzzy multi-criteria decision making model for selecting low-carbon building measures. Procedia Engineering, 118, 606–613. Christen, B., Kjeldsen, C., Dalgaard, T., & Martin-Ortega, J. (2015). Can fuzzy cognitive mapping help in agricultural policy design and communication? Land Use Policy, 45, 64–75. Cootner, P. H. (1964). The random character of stock market prices. MIT Press. Cowan, K., Daim, T., & Anderson, T. (2010). Exploring the impact of technology development and adoption for sustainable hydroelectric power and storage technologies in the Pacific Northwest United States. Energy, 35(12), 4771–4779. Cristóbal, S., & Ramón, J. (2012). A goal programming model for environmental policy analysis: Application to Spain. Energy Policy, 43, 303–307. Crowe, K. A., & Parker, W. H. (2008). Using portfolio theory to guide reforestation and restoration under climate change scenarios. Climatic Change, 89(3), 355–370. Cucchiella, F., Gastaldi, M., & Trosini, M. (2017). Investments and cleaner energy production: A portfolio analysis in the Italian electricity market. Journal of Cleaner Production, 142, 121–132. Cutz, L., Haro, P., Santana, D., & Johnsson, F. (2016). Assessment of biomass energy sources and technologies: The case of Central America. Renewable and Sustainable Energy Reviews, 58, 1411–1431. Dace, E., & Blumberga, D. (2016). How do 28 European Union Member States perform in agricultural greenhouse gas emissions? It depends on what we look at: Application of the multi-criteria analysis. Ecological Indicators, 71, 352–358. de Bremond, A., & Engle, N. L. (2014). Adaptation policies to increase terrestrial ecosystem resilience: Potential utility of a multicriteria approach. Mitigation and Adaptation Strategies for Global Change, 19(3), 331–354. de Bruin, K., Dellink, R. B., Ruijs, A., Bolwidt, L., van Buuren, A., Graveland, J., et al. (2009). Adapting to climate change in The Netherlands: An inventory of climate adaptation options and ranking of alternatives. Climatic Change, 95(1), 23–45. Delarue, E., De Jonghe, C., Belmans, R., & D’haeseleer, W. (2011). Applying portfolio theory to the electricity sector: Energy versus power. Energy Economics, 33(1), 12–23. Den Hartog, H., & Tjan, H. S. (1980). A clay–clay vintage model approach for sectors of industry in Netherlands. De Economist, 128(2), 129–188. Di Lullo, G., Zhang, H., & Kumar, A. (2016). Evaluation of uncertainty in the well-to– tank and combustion greenhouse gas emissions of various transportation fuels. Applied Energy, 184, 413–426. Diakoulaki, D., Antunes, C. H., & Gomes Martins, A. (2005). MCDA and energy planning. Multiple criteria decision analysis: State of the art surveys (pp. 859–890). New York, NY: Springer. Diakoulaki, D., Georgiou, P., Tourkolias, C., Georgopoulou, E., Lalas, D., Mirasgedis, S., et al. (2007). A multicriteria approach to identify investment opportunities for the exploitation of the clean development mechanism. Energy Policy, 35(2), 1088–1099. Dickerson, J. A., & Kosko, B. (1994). Virtual worlds as fuzzy cognitive maps. Presence: Teleoperators & Virtual Environments, 3(2), 173–189. Dixit, A. K., & Pindyck, R. S. (1994). Investment under uncertainty. Princeton University Press. Doukas, H. (2013). Modelling of linguistic variables in multicriteria energy policy support. European Journal of Operational Research, 227(2), 227–238. Doukas, H., Patlitzianas, K. D., & Psarras, J. (2006). Supporting sustainable electricity technologies in Greece using MCDM. Resources Policy, 31(2), 129–136. Doukas, H., Nikas, A., González-Eguino, M., Arto, I., & Anger-Kraavi, A. (2018). From integrated to integrative: Delivering on the Paris agreement. Sustainability, 10(7), 2299. Doumpos, M., & Zopounidis, C. (2011). Preference disaggregation and statistical learning for multicriteria decision support: A review. European Journal of Operational Research, 209(3), 203–214. Dowlatabadi, H. (1995). Integrated assessment models of climate change: An incomplete overview. Energy Policy, 23(4), 289–296. Dowlatabadi, H. (1998). Sensitivity of climate change mitigation estimates to assumptions about technical change. Energy Economics, 20(5), 473–493. Dowlatabadi, H. (20 0 0). Bumping against a gas ceiling. Climatic Change, 46(3), 391–407.

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24 Durbach, I. N., & Stewart, T. J. (2012). Modeling uncertainty in multi-criteria decision analysis. European Journal of Operational Research, 223(1), 1–14. Eden, C., & Ackermann, F. (2013). Making strategy: The journey of strategic management. London: Sage. Edenhofer, O., Pichs-Madruga, R., Sokona, Y., Minx, C. J., Farahani, E., Kadner, S., et al. (2014). Working Group III contribution to the fifth assessment report of the intergovernmental panel on climate change. Intergovernmental Panel on Climate Change. Edmonds, J., & Reiley, J. M. (1985). Global energy-assessing the future. Oxford University Press. Edsand, H. E. (2017). Identifying barriers to wind energy diffusion in Colombia: A function analysis of the technological innovation system and the wider context. Technology in Society, 49, 1–15. Fishbone, L. G., & Abilock, H. (1981). Markal, a linear-programming model for energy systems analysis: Technical description of the bnl version. International Journal of Energy Research, 5(4), 353–375. Flues, F., Löschel, A., Lutz, B. J., & Schenker, O. (2014). Designing an EU energy and climate policy portfolio for 2030: Implications of overlapping regulation under different levels of electricity demand. Energy Policy, 75, 91–99. Forouli, A., Gkonis, N., Nikas, A., Siskos, E., Doukas, H., & Tourkolias, C. (2019). Energy efficiency promotion in Greece in light of risk: Evaluating policies as portfolio assets. Energy, 170, 818–831. Fozer, D., Sziraky, F. Z., Racz, L., Nagy, T., Tarjani, A. J., Toth, A. J., . . . Mizsey, P. (2017). Life cycle, PESTLE and multi-criteria decision analysis of CCS process alternatives. Journal of Cleaner Production, 147, 75–85. Fuss, S., Szolgayová, J., Khabarov, N., & Obersteiner, M. (2012). Renewables and climate change mitigation: Irreversible energy investment under uncertainty and portfolio effects. Energy Policy, 40, 59–68. Füssel, H. (2010). Modeling impacts and adaptation in global IAMs. Wiley Interdisciplinary Reviews: Climate Change, 1(2), 288–303. Geels, F. W. (2012). A socio-technical analysis of low-carbon transitions: introducing the multi-level perspective into transport studies. Journal of Transport Geography, 24, 471–482. Georgakellos, D. A. (2012). Climate change external cost appraisal of electricity generation systems from a life cycle perspective: The case of Greece. Journal of Cleaner production, 32, 124–140. Georgopoulou, E., Sarafidis, Y., Mirasgedis, S., Zaimi, S., & Lalas, D. P. (2003). A multiple criteria decision-aid approach in defining national priorities for greenhouse gases emissions reduction in the energy sector. European Journal of Operational Research, 146(1), 199–215. Ghaderi, S. F., Azadeh, A., Nokhandan, B. P., & Fathi, E. (2012). Behavioral simulation and optimization of generation companies in electricity markets by fuzzy cognitive map. Expert Systems with Applications, 39(5), 4635–4646. Ghafghazi, S., Sowlati, T., Sokhansanj, S., & Melin, S. (2010). A multicriteria approach to evaluate district heating system options. Applied Energy, 87(4), 1134– 1140. Giordano, R., Passarella, G., & Vurro, M. (2010). Fuzzy cognitive maps for conflict analysis and dissolution in drought risk management. Plurimondi: 7. Goulder, L. H. (1995). Environmental taxation and the double dividend: A reader’s guide. International Tax and Public Finance, 2(2), 157–183. Govindan, K., & Jepsen, M. B. (2016). ELECTRE: A comprehensive literature review on methodologies and applications. European Journal of Operational Research, 250(1), 1–29. Gray, S. A., Gray, S., Cox, L. J., & Henly-Shepard, S. (2013). Mental modeler: A fuzzy– logic cognitive mapping modeling tool for adaptive environmental management. In Proceedings of the forty-sixth Hawaii international conference on system sciences (HICSS) (pp. 965–973). IEEE. Gray, S. A., Gray, S., De Kok, J. L., Helfgott, A. E. R., O’Dwyer, B., Jordan, R., et al. (2015). Using fuzzy cognitive mapping as a participatory approach to analyze change, preferred states, and perceived resilience of social-ecological systems. Ecology and Society, 20(2). Gray, S. R. J, Gagnon, A. S., Gray, S. A., O’Dwyer, B., O’Mahony, C., Muir, D., et al. (2014). Are coastal managers detecting the problem? Assessing stakeholder perception of climate vulnerability using Fuzzy Cognitive Mapping. Ocean & Coastal Management, 94, 74–89. Greco, S., Ehrgott, M., & Figueira, J. R. (2016). Multiple criteria decision analysis: State of the art surveys. (Eds). International Series in Operations Research & Management Science: 1–2 (second ed.). Springer. Hedenus, F., Azar, C., & Lindgren, K. (2006). Induced technological change in a limited foresight optimization model. The Energy Journal, 27, 109–122. Hellsmark, H., & Jacobsson, S. (2009). Opportunities for and limits to academics as system builders—The case of realizing the potential of gasified biomass in Austria. Energy Policy, 37(12), 5597–5611. Heo, E., Kim, J., & Boo, K. J. (2010). Analysis of the assessment factors for renewable energy dissemination program evaluation using fuzzy AHP. Renewable and Sustainable Energy Reviews, 14(8), 2214–2220. Hobbs, B. F., Ludsin, S. A., Knight, R. L., Ryan, P. A., Biberhofer, J., & Ciborowski, J. J. (2002). Fuzzy cognitive mapping as a tool to define management objectives for complex ecosystems. Ecological Applications, 12(5), 1548–1565. Hope, C. (2006). The marginal impact of CO2 from PAGE2002: An integrated assessment model incorporating the IPCC’s five reasons for concern. Integrated Assessment, 6(1), 19–56. Hsueh, S. L. (2015). Assessing the effectiveness of community-promoted environmental protection policy by using a Delphi-fuzzy method: A case study on solar power and plain afforestation in Taiwan. Renewable and Sustainable Energy Reviews, 49, 1286–1295.

21

Hua, S., Liang, J., Zeng, G., Xu, M., Zhang, C., Yuan, Y., et al. (2015). How to manage future groundwater resource of China under climate change and urbanization: An optimal stage investment design from modern portfolio theory. Water Research, 85, 31–37. Huang, S. C., Lo, S. L., & Lin, Y. C. (2013). Application of a fuzzy cognitive map based on a structural equation model for the identification of limitations to the development of wind power. Energy Policy, 63, 851–861. Huang, Y. H., & Wu, J. H. (2008). A portfolio risk analysis on electricity supply planning. Energy Policy, 36(2), 627–641. Huff, A. S. (1990). Mapping strategic thought. John Wiley & Sons. Humpenöder, F., Schaldach, R., Cikovani, Y., & Schebek, L. (2013). Effects of land-use change on the carbon balance of 1st generation biofuels: An analysis for the European Union combining spatial modeling and LCA. Biomass and Bioenergy, 56, 166–178. International Emissions Trading Association (2006). Wien Automatic System Planning (WASP) package: A computer code for power generating system expansion planning version WASP-IV with user interface user’s manual (pp. 13–150). Vienna, Austria: IAEA. Jansen, J. C., Beurskens, L., & Van Tilburg, X. (2006). Application of portfolio analysis to the Dutch generating mix. ECN Energy Research Center at the Netherlands (ECN) Report C-05-100. Javid, R. J., Nejat, A., & Hayhoe, K. (2014). Selection of CO2 mitigation strategies for road transportation in the United States using a multi-criteria approach. Renewable and Sustainable Energy Reviews, 38, 960–972. Jayaraman, R., Colapinto, C., La Torre, D., & Malik, T. (2015). Multi-criteria model for sustainable development using goal programming applied to the United Arab Emirates. Energy Policy, 87, 447–454. Jebaraj, S., & Iniyan, S. (2006). A review of energy models. Renewable and Sustainable Energy Reviews, 10(4), 281–311. Jetter, A., & Schweinfort, W. (2011). Building scenarios with fuzzy cognitive maps: An exploratory study of solar energy. Futures, 43(1), 52–66. Jun, K. S., Chung, E. S., Kim, Y. G., & Kim, Y. (2013). A fuzzy multi-criteria approach to flood risk vulnerability in South Korea by considering climate change impacts. Expert Systems with Applications, 40(4), 1003–1013. Kafetzis, A., McRoberts, N., & Mouratiadou, I. (2010). Using fuzzy cognitive maps to support the analysis of stakeholders’ views of water resource use and water quality policy. Fuzzy cognitive maps (pp. 383–402). Berlin Heidelberg: Springer. Karakosta, C., Doukas, H., & Psarras, J. (2009). Directing clean development mechanism towards developing countries’ sustainable development priorities. Energy for Sustainable Development, 13(2), 77–84. Karavas, C. S., Kyriakarakos, G., Arvanitis, K. G., & Papadakis, G. (2015). A multi-agent decentralized energy management system based on distributed intelligence for the design and control of autonomous polygeneration microgrids. Energy Conversion and Management, 103, 166–179. Kaya, T., & Kahraman, C. (2011). Multicriteria decision making in energy planning using a modified fuzzy TOPSIS methodology. Expert Systems with Applications, 38(6), 6577–6585. Kayikci, Y., & Stix, V. (2014). Causal mechanism in transport collaboration. Expert Systems with Applications, 41(4), 1561–1575. Kelly, D. L., & Kolstad, C. D. (1999). Integrated assessment models for climate change control. International Yearbook of Environmental and Resource Economics, 2000, 171–197. Klein, S. J., & Whalley, S. (2015). Comparing the sustainability of US electricity options through multi-criteria decision analysis. Energy Policy, 79, 127–149. Konidari, P., & Mavrakis, D. (2007). A multi-criteria evaluation method for climate change mitigation policy instruments. Energy Policy, 35(12), 6235–6257. Kontogianni, A., Papageorgiou, E., Salomatina, L., Skourtos, M., & Zanou, B. (2012). Risks for the Black Sea marine environment as perceived by Ukrainian stakeholders: A fuzzy cognitive mapping application. Ocean & Coastal Management, 62, 34–42. Kontogianni, A., Tourkolias, C., & Papageorgiou, E. I. (2013). Revealing market adaptation to a low carbon transport economy: Tales of hydrogen futures as perceived by fuzzy cognitive mapping. International Journal of Hydrogen Energy, 38(2), 709–722. Kosko, B. (1986). Fuzzy cognitive maps. International Journal of Man–Machine Studies, 24(1), 65–75. Kottas, T. L., Boutalis, Y. S., & Karlis, A. D. (2006). New maximum power point tracker for PV arrays using fuzzy controller in close cooperation with fuzzy cognitive networks. IEEE Transactions on Energy Conversion, 21(3), 793–803. Kratena, K., & Streicher, G. (2009). Macroeconomic input-output modelling: Structures, functional forms and closure rules. Austrian Institute of Economic Research. International Input–Output Association Working Paper WPIOX, 09-009. Krohling, R. A., & Campanharo, V. C. (2011). Fuzzy TOPSIS for group decision making: A case study for accidents with oil spill in the sea. Expert Systems with Applications, 38(4), 4190–4197. Kurosawa, A., Yagita, H., Zhou, W., Tokimatsu, K., & Yanagisawa, Y. (1999). Analysis of carbon emission stabilization targets and adaptation by integrated assessment model. The Energy Journal, 20, 157–175. Kyriakarakos, G., Dounis, A. I., Arvanitis, K. G., & Papadakis, G. (2012). A fuzzy cognitive maps–petri nets energy management system for autonomous polygeneration microgrids. Applied Soft Computing, 12(12), 3785–3797. Kyriakarakos, G., Patlitzianas, K., Damasiotis, M., & Papastefanakis, D. (2014). A fuzzy cognitive maps decision support system for renewables local planning. Renewable and Sustainable Energy Reviews, 39, 209–222. Lai, X., Ye, Z., Xu, Z., Holmes, M. H., & Lambright, W. H. (2012). Carbon capture and sequestration (CCS) technological innovation system in China: Structure, function evaluation and policy implication. Energy Policy, 50, 635–646.

22

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Lai, Y. J., Liu, T. Y., & Hwang, C. L. (1994). Topsis for MODM. European Journal of Operational Research, 76(3), 486–500. Lange, A., & Treich, N. (2008). Uncertainty, learning and ambiguity in economic models on climate policy: Some classical results and new directions. Climatic Change, 89(1), 7–21. Laurikka, H., & Springer, U. (2003). Risk and return of project-based climate change mitigation: A portfolio approach. Global Environmental Change, 13(3), 207–217. Le Téno, J. F., & Mareschal, B. (1998). An interval version of PROMETHEE for the comparison of building products’ design with ill-defined data on environmental quality. European Journal of Operational Research, 109(2), 522–529. Lejour, A., Veenendaal, P., Verweij, G., & van Leeuwen, N. (2006). WorldScan: A model for international economic policy analysis (No. 111). CPB Netherlands Bureau for Economic Policy Analysis. Lemoine, D. M., Fuss, S., Szolgayova, J., Obersteiner, M., & Kammen, D. M. (2012). The influence of negative emission technologies and technology policies on the optimal climate mitigation portfolio. Climatic Change, 113(2), 141–162. Lintner, J. (1975). The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets. Stochastic Optimization Models in Finance (pp. 131–155). Elsevier. Lintunen, J., & Uusivuori, J. (2016). On the economics of forests and climate change: Deriving optimal policies. Journal of Forest Economics, 24, 130–156. Liu, L. (1999). Approximate portfolio analysis. European Journal of Operational Research, 119(1), 35–49. Liu, M., & Wu, F. F. (2007). Portfolio optimization in electricity markets. Electric Power Systems Research, 77(8), 10 0 0–10 09. Lopolito, A., Nardone, G., Prosperi, M., Sisto, R., & Stasi, A. (2011). Modeling the bio-refinery industry in rural areas: A participatory approach for policy options comparison. Ecological Economics, 72, 18–27. Loulou, R., Remme, U., Kanudia, A., Lehtila, A., & Goldstein, G. (2005). Documentation for the TIMES Model Part II. In Proceedings of the energy technology systems analysis programme (ETSAP). Luo, C., & Wu, D. (2016). Environment and economic risk: An analysis of carbon emission market and portfolio management. Environmental Research, 149, 297–301. Luthra, S., Mangla, S. K., & Kharb, R. K. (2015). Sustainable assessment in energy planning and management in Indian perspective. Renewable and Sustainable Energy Reviews, 47, 58–73. Madlener, R., Antunes, C. H., & Dias, L. C. (2009). Assessing the performance of biogas plants with multi-criteria and data envelopment analysis. European Journal of Operational Research, 197(3), 1084–1094. Maimoun, M., Madani, K., & Reinhart, D. (2016). Multi-level multi-criteria analysis of alternative fuels for waste collection vehicles in the United States. Science of the Total Environment, 550, 349–361. Mallampalli, V. R., Mavrommati, G., Thompson, J., Duveneck, M., Meyer, S., et al. (2016). Methods for translating narrative scenarios into quantitative assessments of land use change. Environmental Modelling & Software, 82, 7–20. Manne, A. S., & Richels, R. G. (2005). MERGE: An integrated assessment model for global climate change. Energy and environment (pp. 175–189). Springer. Marinoni, O., Adkins, P., & Hajkowicz, S. (2011). Water planning in a changing climate: Joint application of cost utility analysis and modern portfolio theory. Environmental Modelling & Software, 26(1), 18–29. Markowitz, H. (1952). Portfolio selection. The Journal of Finance, 7(1), 77–91. Marrero, G. A., Puch, L. A., & Ramos-Real, F. J. (2015). Mean-variance portfolio methods for energy policy risk management. International Review of Economics & Finance, 40, 246–264. Marttunen, M., Lienert, J., & Belton, V. (2017). Structuring problems for multi-criteria decision analysis in practice: A literature review of method combinations. European Journal of Operational Research, 263(1), 1–17. Masui, T., Hanaoka, T., Hikita, S., & Kainuma, M. (2006). Assessment of CO3 reductions and economic impacts considering energy-saving investments. The Energy Journal, 175–190. Mattsson, N., & Wene, C. O. (1997). Assessing new energy technologies using an energy system model with endogenized experience curves. International Journal of Energy Research, 21(4), 385–393. Mavrotas, G., & Florios, K. (2013). An improved version of the augmented ε -constraint method (AUGMECON2) for finding the exact Pareto set in multi-objective integer programming problems. Applied Mathematics and Computation, 219(18), 9652–9669. Mavrotas, G., Figueira, J. R., & Siskos, E. (2015). Robustness analysis methodology for multi-objective combinatorial optimization problems and application to project selection. Omega, 52, 142–155. McLohglin, E., & Bazilian, M. (2006). Application of portfolio analysis to the irish generating mix in 2020. Sustainable Energy Ireland (SEI). Meliadou, A., Santoro, F., Nader, M. R., Dagher, M. A., Al Indary, S., & Salloum, B. A. (2012). Prioritising coastal zone management issues through fuzzy cognitive mapping approach. Journal of Environmental Management, 97, 56–68. Mendelsohn, R., & Seo, S. N. (2007). Climate change adaptation in Africa: a microeconomic analysis of livestock choice. The World Bank. Messner, S. (1997). Endogenized technological learning in an energy systems model. Journal of Evolutionary Economics, 7(3), 291–313. Michailidou, A. V., Vlachokostas, C., & Moussiopoulos, N (2016). Interactions between climate change and the tourism sector: Multiple-criteria decision analysis to assess mitigation and adaptation options in tourism areas. Tourism Management, 55, 1–12. Miettinen, P., & Hämäläinen, R. P. (1997). How to benefit from decision analysis in environmental life cycle assessment (LCA). European Journal of Operational Research, 102(2), 279–294.

Miller, K. A., & Belton, V. (2014). Water resource management and climate change adaptation: A holistic and multiple criteria perspective. Mitigation and Adaptation Strategies for Global Change, 19(3), 289–308. Ministry of the Environment and Energy (2017). 4th National Energy Efficiency Action Plan of Greece. Available at: https://ec.europa.eu/energy/sites/ener/files/ documents/el_neeap_2017_en.pdf. Mitter, H., Heumesser, C., & Schmid, E. (2015). Spatial modeling of robust crop production portfolios to assess agricultural vulnerability and adaptation to climate change. Land Use Policy, 46, 75–90. Moallemi, E. A., de Haan, F. J., Webb, J. M., George, B. A., & Aye, L. (2017). Transition dynamics in state-influenced niche empowerments: Experiences from India’s electricity sector. Technological Forecasting and Social Change, 116, 129– 141. Mohamadabadi, H. S., Tichkowsky, G., & Kumar, A. (2009). Development of a multi-criteria assessment model for ranking of renewable and non-renewable transportation fuel vehicles. Energy, 34(1), 112–125. Montanari, R. (2004). Environmental efficiency analysis for enel thermo-power plants. Journal of Cleaner Production, 12(4), 403–414. Mourhir, A., Rachidi, T., Papageorgiou, E. I., Karim, M., & Alaoui, F. S. (2016). A cognitive map framework to support integrated environmental assessment. Environmental Modelling & Software, 77, 81–94. Munda, G. (2005). Multiple criteria decision analysis and sustainable development. Multiple criteria decision analysis: State of the art surveys (pp. 953–986). New York: Springer. Mundaca, L., Neij, L., Worrell, E., & McNeil, M. (2010). Evaluating energy efficiency policies with energy-economy models. Annual Review of Environment and Resources, 35, 305–344. Muñoz, J. I., de la Nieta, A. A. S., Contreras, J., & Bernal-Agustín, J. L. (2009). Optimal investment portfolio in renewable energy: The Spanish case. Energy Policy, 37(12), 5273–5284. Narita, D., & Klepper, G. (2016). Economic incentives for carbon dioxide storage under uncertainty: A real options analysis. International Journal of Greenhouse Gas Control, 53, 18–27. Natarajan, R., Subramanian, J., & Papageorgiou, E. I. (2016). Hybrid learning of fuzzy cognitive maps for sugarcane yield classification. Computers and Electronics in Agriculture, 127, 147–157. Nazari, M. S., Maybee, B., Whale, J., & McHugh, A. (2015). Climate policy uncertainty and power generation investments: A real options-CVaR portfolio optimization approach. Energy Procedia, 75, 2649–2657. Neves, L. P., Martins, A. G., Antunes, C. H., & Dias, L. C. (2008). A multi-criteria decision approach to sorting actions for promoting energy efficiency. Energy Policy, 36(7), 2351–2363. Nikas, A., Gkonis, N., Forouli, A., Siskos, E., Arsenopoulos, A., Papapostolou, A., et al. (2019). Greece: From near-term actions to long-term pathways: Risks and uncertainties associated with the national energy efficiency framework. In S. Hanger-Kopp, J. Lieu, & A. Nikas (Eds.), Narratives of low-carbon transitions: Understanding risks and uncertainties. Abingdon: Routledge. Nikas, A., & Doukas, H. (2016). Developing robust climate policies: A fuzzy cognitive map approach. Robustness analysis in decision aiding, optimization, and analytics (pp. 239–263). Springer International Publishing. Nikas, A., Doukas, H., Lieu, J., Alvarez Tinoco, R., Charisopoulos, V., Charisopoulos, V., et al. (2017). Managing stakeholder knowledge for the evaluation of innovation systems in the face of climate change. Journal of Knowledge Management, 21(5), 1013–1034. Nikas, A., Ntanos, E., & Doukas, H. (2019). A semi-quantitative modelling application for assessing energy efficiency strategies. Applied Soft Computing, 76, 140–155. Nikas, A., Stavrakas, V., Arsenopoulos, A., Doukas, H., Antosiewicz, M., WitajewskiBaltvilks, J., et al. (2019). Barriers to and consequences of a solar-based energy transition in Greece. Environmental Innovation and Societal Transitions in press. doi:10.1016/j.eist.2018.12.004. Nordhaus, W. D. (1994). Managing the global commons: The economics of climate change: Vol. 31. Cambridge, MA: MIT press. Nordhaus, W. D. (2008). A question of balance: Weighing the options on global warming policies. Yale University Press. O’Neill, B. C., Kriegler, E., Ebi, K. L., Kemp-Benedict, E., Riahi, K., Rothman, D. S., et al. (2017). The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century. Global Environmental Change, 42, 169–180. O’Neill, B. C., Kriegler, E., Riahi, K., Ebi, K. L., Hallegatte, S., Carter, T. R., et al. (2014). A new scenario framework for climate change research: The concept of shared socioeconomic pathways. Climatic Change, 122(3), 387–400. Oda, J., & Akimoto, K. (2011). An analysis of CCS investment under uncertainty. Energy Procedia, 4, 1997–2004. Olazabal, M., & Pascual, U. (2016). Use of fuzzy cognitive maps to study urban resilience and transformation. Environmental Innovation and Societal Transitions, 18, 18–40. Oliveira, C., & Antunes, C. H. (2004). A multiple objective model to deal with economy–energy–environment interactions. European Journal of Operational Research, 153(2), 370–385. Onar, S. C., Oztaysi, B., Otay, I.˙ , & Kahraman, C. (2015). Multi-expert wind energy technology selection using interval-valued intuitionistic fuzzy sets. Energy, 90, 274–285. Onu, P. U., Quan, X., Xu, L., Orji, J., & Onu, E. (2017). Evaluation of sustainable acid rain control options utilizing a fuzzy TOPSIS multi-criteria decision analysis model frame work. Journal of Cleaner Production, 141, 612–625.

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24 Opricovic, S., & Tzeng, G. H. (2004). Compromise solution by MCDM methods: A comparative analysis of VIKOR and TOPSIS. European Journal of Operational Research, 156(2), 445–455. Ortiz, R., & Markandya, A. (2010). Literature review of integrated impact assessment models of climate change with emphasis on damage functions. Barcelona, Spain: BC3 Working paper series 2009-06. Ortolani, L., McRoberts, N., Dendoncker, N., & Rounsevell, M. (2010). Analysis of farmers’ concepts of environmental management measures: An application of cognitive maps and cluster analysis in pursuit of modelling agents’ behaviour. Fuzzy cognitive maps (pp. 363–381). Berlin Heidelberg: Springer. Özesmi, U. (2006a). Ecosystems in the mind: Fuzzy cognitive maps of the Kizilirmak Delta Wetlands in Turkey. arXiv preprint q-bio/0603022. Özesmi, U. (2006b). Fuzzy cognitive maps of local people impacted by dam construction: Their demands regarding resettlement. arXiv preprint q-bio/0601032. Özesmi, U., & Özesmi, S. (2003). A participatory approach to ecosystem conservation: Fuzzy cognitive maps and stakeholder group analysis in Uluabat Lake, Turkey. Environmental Management, 31(4), 518–531. Papadopoulos, A., & Karagiannidis, A. (2008). Application of the multi-criteria analysis method Electre III for the optimisation of decentralised energy systems. Omega, 36(5), 766–776. Papageorgiou, E. I., Markinos, A. T., & Gemtos, T. A. (2011). Fuzzy cognitive map based approach for predicting yield in cotton crop production as a basis for decision support system in precision agriculture application. Applied Soft Computing, 11(4), 3643–3657. Papageorgiou, E., & Kontogianni, A. (2012). Using fuzzy cognitive mapping in environmental decision making and management: A methodological primer and an application. INTECH Open Access Publisher. Parrado, R., & De Cian, E. (2014). Technology spillovers embodied in international trade: Intertemporal, regional and sectoral effects in a global CGE framework. Energy Economics, 41, 76–89. Parson, E. A., & Fisher-Vanden, A. K. (1997). Integrated assessment models of global climate change. Annual Review of Energy and the Environment, 22(1), 589–628. Paul, S., Sarkar, B., & Bose, P. K. (2015). Eclectic decision for the selection of tree borne oil (TBO) as alternative fuel for internal combustion engine. Renewable and Sustainable Energy Reviews, 48, 256–263. Peng, Z., Wu, L., & Chen, Z. (2016). Research on steady states of fuzzy cognitive map and its application in three-rivers ecosystem. Sustainability, 8(1), 40. Perkoulidis, G., Papageorgiou, A., Karagiannidis, A., & Kalogirou, S. (2010). Integrated assessment of a new Waste-to-Energy facility in Central Greece in the context of regional perspectives. Waste Management, 30(7), 1395–1406. Phillips, L. D., & Bana e Costa, C. A. (2007). Transparent prioritisation, budgeting and resource allocation with multi-criteria decision analysis and decision conferencing. Annals of Operations Research, 154(1), 51–68. Pilavachi, P. A., Stephanidis, S. D., Pappas, V. A., & Afgan, N. H. (2009). Multi-criteria evaluation of hydrogen and natural gas fuelled power plant technologies. Applied Thermal Engineering, 29(11), 2228–2234. Promentilla, M. A. B., Aviso, K. B., & Tan, R. R. (2014). A group fuzzy analytic network process to prioritize low carbon energy systems in the Philippines. Energy Procedia, 61, 808–811. Pugh, G., Clarke, L., Marlay, R., Kyle, P., Wise, M., McJeon, H., et al. (2011). Energy R&D portfolio analysis based on climate change mitigation. Energy Economics, 33(4), 634–643. Rajaram, T., & Das, A. (2010). Modeling of interactions among sustainability components of an agro-ecosystem using local knowledge through cognitive mapping and fuzzy inference system. Expert Systems with Applications, 37(2), 1734– 1744. Ramazankhani, M. E., Mostafaeipour, A., Hosseininasab, H., & Fakhrzad, M. B. (2016). Feasibility of geothermal power assisted hydrogen production in Iran. International Journal of Hydrogen Energy, 41(41), 18351–18369. Rana, A., & Morita, T. (20 0 0). Scenarios for greenhouse gas emission mitigation: A review of modeling of strategies and policies in integrated assessment models. Environmental Economics and Policy Studies, 3(2), 267–289. Reckien, D. (2014). Weather extremes and street life in India—Implications of Fuzzy Cognitive Mapping as a new tool for semi-quantitative impact assessment and ranking of adaptation measures. Global Environmental Change, 26, 1–13. Ren, J., & Lützen, M. (2015). Fuzzy multi-criteria decision-making method for technology selection for emissions reduction from shipping under uncertainties. Transportation Research Part D: Transport and Environment, 40, 43–60. Ribeiro, F., Ferreira, P., & Araújo, M. (2013). Evaluating future scenarios for the power generation sector using a Multi-Criteria Decision Analysis (MCDA) tool: The Portuguese case. Energy, 52, 126–136. Rojas-Zerpa, J. C., & Yusta, J. M. (2015). Application of multicriteria decision methods for electric supply planning in rural and remote areas. Renewable and Sustainable Energy Reviews, 52, 557–571. Romejko, K., & Nakano, M. (2017). Portfolio analysis of alternative fuel vehicles considering technological advancement, energy security and policy. Journal of Cleaner Production, 142, 39–49. Roques, F. A., Newbery, D. M., & Nuttall, W. J. (2008). Fuel mix diversification incentives in liberalized electricity markets: A mean–variance portfolio theory approach. Energy Economics, 30(4), 1831–1849. Roth, S., Hirschberg, S., Bauer, C., Burgherr, P., Dones, R., Heck, T., et al. (2009). Sustainability of electricity supply technology portfolio. Annals of Nuclear Energy, 36(3), 409–416. Roy, B. (1985). Méthodologie multicritère d" aide à la décision. Paris: Economica. Roy, B. (1990). Decision-aid and decision-making. European Journal of Operational Research, 45(2–3), 324–331.

23

Roy, B., & Vanderpooten, D. (1997). An overview on “The European school of MCDA: Emergence, basic features and current works”. European Journal of Operational Research, 99(1), 26–27. Roy, B., Présent, D. M., & Silhol, D. (1986). A programming method for determining which Paris metro stations should be renovated. European Journal of Operational Research, 24(2), 318–334. Røyne, F., Penaloza, D., Sandin, G., Berlin, J., & Svanström, M. (2016). Climate impact assessment in life cycle assessments of forest products: Implications of method choice for results and decision-making. Journal of Cleaner Production, 116, 90–99. Saaty, T. L. (1990). How to make a decision: The analytic hierarchy process. European Journal of Operational Research, 48(1), 9–26. Sacchelli, S. (2014). Social acceptance optimization of biomass plants: A fuzzy cognitive map and evolutionary algorithm application. Chemical Engineering, 37, 181–186. Sadeghi, A., Larimian, T., & Molabashi, A. (2012). Evaluation of renewable energy sources for generating electricity in province of Yazd: A fuzzy MCDM approach. Procedia-Social and Behavioral Sciences, 62, 1095–1099. Sakthivel, G., Ilangkumaran, M., & Gaikwad, A. (2015). A hybrid multi-criteria decision modeling approach for the best biodiesel blend selection based on ANP– TOPSIS analysis. Ain Shams Engineering Journal, 6(1), 239–256. Samarasinghe, S., & Strickert, G. (2013). Mixed-method integration and advances in fuzzy cognitive maps for computational policy simulations for natural hazard mitigation. Environmental Modelling & Software, 39, 188–200. Sano, F., Akimoto, K., Homma, T., & Tomoda, T. (2006). Analysis of technological portfolios for CO2 stabilizations and effects of technological changes. The Energy Journal, 27, 141–161. Santos-Alamillos, F. J., Thomaidis, N. S., Usaola-García, J., Ruiz-Arias, J. A., & Pozo-Vázquez, D. (2017). Exploring the mean-variance portfolio optimization approach for planning wind repowering actions in Spain. Renewable Energy, 106, 335–342. Schneider, S. H. (1997). Integrated assessment modeling of global climate change: Transparent rational tool for policy making or opaque screen hiding value-laden assumptions? Environmental Modeling and Assessment, 2(4), 229–249. Schwanitz, V. J. (2013). Evaluating integrated assessment models of global climate change. Environmental Modelling & Software, 50, 120–131. Scrieciu, S. S¸ ., & Chalabi, Z. (2014). Climate policy planning and development impact assessment. Mitigation and Adaptation Strategies for Global Change, 19(3), 255–260. Scrieciu, S. S¸ ., Belton, V., Chalabi, Z., Mechler, R., & Puig, D. (2014). Advancing methodological thinking and practice for development-compatible climate policy planning. Mitigation and Adaptation Strategies for Global Change, 19(3), 261–288. S¸ engül, Ü., Eren, M., Shiraz, S. E., Gezder, V., & S¸ engül, A. B. (2015). Fuzzy TOPSIS method for ranking renewable energy supply systems in Turkey. Renewable Energy, 75, 617–625. Shahnazari, M., McHugh, A., Maybee, B., & Whale, J. (2017). Overlapping carbon pricing and renewable support schemes under political uncertainty: Global lessons from an Australian case study. Applied Energy, 200, 237–248. Shakouri, M., Lee, H. W., & Choi, K. (2015). PACPIM: New decision-support model of optimized portfolio analysis for community-based photovoltaic investment. Applied Energy, 156, 607–617. Sharpe, W. F. (1964). Capital asset prices: A theory of market equilibrium under conditions of risk. The Journal of Finance, 19(3), 425–442. Shiau, T. A., & Liu, J. S. (2013). Developing an indicator system for local governments to evaluate transport sustainability strategies. Ecological Indicators, 34, 361–371. Shmelev, S. E., & van den Bergh, J. C. (2016). Optimal diversity of renewable energy alternatives under multiple criteria: An application to the UK. Renewable and Sustainable Energy Reviews, 60, 679–691. Shreve, C. M., & Kelman, I. (2014). Does mitigation save? Reviewing cost-benefit analyses of disaster risk reduction. International Journal of Disaster Risk Reduction, 10, 213–235. Siddiqui, A. S., Tanaka, M., & Chen, Y. (2016). Are targets for renewable portfolio standards too low? The impact of market structure on energy policy. European Journal of Operational Research, 250(1), 328–341. Singh, P. K., & Nair, A. (2014). Livelihood vulnerability assessment to climate variability and change using fuzzy cognitive mapping approach. Climatic Change, 127(3–4), 475–491. Siskos, E., & Tsotsolas, N. (2015). Elicitation of criteria importance weights through the Simos method: A robustness concern. European Journal of Operational Research, 246(2), 543–553. Siskos, Y., Grigoroudis, E., & Matsatsinis, N. F. (2005). UTA methods. Multiple criteria decision analysis: State of the art surveys (pp. 297–334). New York: Springer. Soderholm, P. (2007). Modelling the economic costs of climate policy. Lulea University of Technology. Soler, L. S., Kok, K., Camara, G., & Veldkamp, A. (2012). Using fuzzy cognitive maps to describe current system dynamics and develop land cover scenarios: A case study in the Brazilian Amazon. Journal of Land Use Science, 7(2), 149–175. Springer, U. (2003). Can the risks of the Kyoto mechanisms be reduced through Portfolio diversification? Evidence from the Swedish AIJ Program. Environmental and Resource Economics, 25(4), 501–513. Stach, W., Kurgan, L., & Pedrycz, W. (2010). Expert-based and computational methods for developing fuzzy cognitive maps. Fuzzy cognitive maps (pp. 23–41). Berlin Heidelberg: Springer. Stamford, L., & Azapagic, A. (2014). Life cycle sustainability assessment of UK electricity scenarios to 2070. Energy for Sustainable Development, 23, 194–211.

24

H. Doukas and A. Nikas / European Journal of Operational Research 280 (2020) 1–24

Stanton, E., Ackerman, F., & Kartha, S. (2009). Inside the integrated assessment models: Four issues in climate economics. Climate and Development, 1(2), 166– 184. Stern, N. H. (2007). The economics of climate change: The stern review. Cambridge University Press. Streimikiene, D., & Baležentis, T. (2013a). Multi-criteria assessment of small scale CHP technologies in buildings. Renewable and Sustainable Energy Reviews, 26, 183–189. Streimikiene, D., & Baležentis, T. (2013b). Multi-objective ranking of climate change mitigation policies and measures in Lithuania. Renewable and Sustainable Energy Reviews, 18, 144–153. ˙ I., & Balezentis, A. (2012). Prioritizing Streimikiene, D., Baležentis, T., Krisciukaitiene, sustainable electricity production technologies: MCDM approach. Renewable and Sustainable Energy Reviews, 16(5), 3302–3311. ˙ D., Šliogeriene, ˙ J., & Turskis, Z. (2016). Multi-criteria analysis of elecŠtreimikiene, tricity generation technologies in Lithuania. Renewable Energy, 85, 148–156. Talaei, A., Ahadi, M. S., & Maghsoudy, S. (2014). Climate friendly technology transfer in the energy sector: A case study of Iran. Energy Policy, 64, 349–363. Theodorou, S., Florides, G., & Tassou, S. (2010). The use of multiple criteria decision making methodologies for the promotion of RES through funding schemes in Cyprus: A review. Energy Policy, 38(12), 7783–7792. Tol, R. S. (1997). On the optimal control of carbon dioxide emissions: An application of FUND. Environmental Modeling and Assessment, 2(3), 151–163. Tol, R. S. (2012). A cost–benefit analysis of the EU 20/20/2020 package. Energy Policy, 49, 288–295. Torani, K., Rausser, G., & Zilberman, D. (2016). Innovation subsidies versus consumer subsidies: A real options analysis of solar energy. Energy Policy, 92, 255–269. Tsai, W. H., Yang, C. H., Chang, J. C., & Lee, H. L. (2014). An activity-based costing decision model for life cycle assessment in green building projects. European Journal of Operational Research, 238(2), 607–619. Tsoutsos, T., Drandaki, M., Frantzeskaki, N., Iosifidis, E., & Kiosses, I. (2009). Sustainable energy planning by using multi-criteria analysis application in the Island of Crete. Energy Policy, 37(5), 1587–1600. Ulutas¸ , B. H. (2005). Determination of the appropriate energy policy for Turkey. Energy, 30(7), 1146–1161. Vahabzadeh, A. H., Asiaei, A., & Zailani, S. (2015). Green decision-making model in reverse logistics using FUZZY-VIKOR method. Resources, Conservation and Recycling, 103, 125–138. Vaillancourt, K., & Waaub, J. P. (2004). Equity in international greenhouse gases abatement scenarios: A multicriteria approach. European Journal of Operational Research, 153(2), 489–505. Van Asseldonk, M. A., & Langeveld, J. W. A. (2007). Coping with climate change in agriculture: A portfolio analysis. 101st Seminar of European Association of Agricultural Economists, Berlin. EAAE: Germany. Van den Bergh, J. C. (2004). Optimal climate policy is a utopia: From quantitative to qualitative cost-benefit analysis. Ecological Economics, 48(4), 385–393. Van den Bergh, J. C. J. M., & Botzen, W. J. W. (2015). Monetary valuation of the social cost of CO2 emissions: A critical survey. Ecological Economics, 114, 33–46. van Vliet, M., Kok, K., & Veldkamp, T. (2010). Linking stakeholders and modellers in scenario studies: The use of fuzzy cognitive maps as a communication and learning tool. Futures, 42(1), 1–14. Vanwindekens, F. M., Stilmant, D., & Baret, P. V. (2013). Development of a broadened cognitive mapping approach for analysing systems of practices in social–ecological systems. Ecological Modelling, 250, 352–362.

Vasslides, J. M., & Jensen, O. P. (2016). Fuzzy cognitive mapping in support of integrated ecosystem assessments: Developing a shared conceptual model among stakeholders. Journal of Environmental Management, 166, 348–356. Voinov, A., & Bousquet, F. (2010). Modelling with stakeholders. Environmental Modelling & Software, 25(11), 1268–1281. Volkart, K., Bauer, C., Burgherr, P., Hirschberg, S., Schenler, W., & Spada, M. (2016). Interdisciplinary assessment of renewable, nuclear and fossil power generation with and without carbon capture and storage in view of the new Swiss energy policy. International Journal of Greenhouse Gas Control, 54, 1–14. Vörös, J. (1986). Portfolio analysis—An analytic derivation of the efficient portfolio frontier. European Journal of Operational Research, 23(3), 294–300. Watkiss, P., Downing, T., & Dyszynski, J. (2010). AdaptCost project: Analysis of the economic costs of climate change adaptation in Africa. Nairobi: UNEP. Wei, Y. M., Mi, Z. F., & Huang, Z. (2015). Climate policy modeling: An online SCI-E and SSCI based literature review. Omega, 57, 70–84. Westner, G., & Madlener, R. (2010). The benefit of regional diversification of cogeneration investments in Europe: A mean-variance portfolio analysis. Energy Policy, 38(12), 7911–7920. White, B. (2007). A mean-variance portfolio optimization of California’s generation mix to 2020: Achieving California’s 33 percent renewable portfolio standard goal. Draft Consultant Report. California Energy Commission. Wildenberg, M., Bachhofer, M., Adamescu, M., De Blust, G., Diaz-Delgadod, R., Isak, K., et al. (2010). Linking thoughts to flows-fuzzy cognitive mapping as tool for integrated landscape modelling. In Proceedings of the 2010 international conference on integrative landscape modeling: Linking environmental, social and computer science: Vol. 3 (p. 5). Worrell, E., Ramesohl, S., & Boyd, G. (2004). Advances in energy forecasting models based on engineering economics. Annual Review of Environment and Resources, 29, 345–381. Xu, B., Nayak, A., Gray, D., & Ouenniche, J. (2016). Assessing energy business cases implemented in the North Sea Region and strategy recommendations. Applied Energy, 172, 360–371. Yap, H. Y., & Nixon, J. D. (2015). A multi-criteria analysis of options for energy recovery from municipal solid waste in India and the UK. Waste Management, 46, 265–277. Zhang, H., Song, J., Su, C., & He, M. (2013). Human attitudes in environmental management: Fuzzy cognitive maps and policy option simulations analysis for a coal-mine ecosystem in China. Journal of Environmental Management, 115, 227–234. Zhao, Z. Y., Zhu, J., & Zuo, J. (2014). Sustainable development of the wind power industry in a complex environment: A flexibility study. Energy Policy, 75, 392–397. Zhou, W., Zhu, B., Fuss, S., Szolgayová, J., Obersteiner, M., & Fei, W. (2010). Uncertainty modeling of CCS investment strategy in China’s power sector. Applied Energy, 87(7), 2392–2400. Zhu, L., & Fan, Y. (2010). Optimization of China’s generating portfolio and policy implications based on portfolio theory. Energy, 35(3), 1391–1402. Zhu, L., & Fan, Y. (2011). A real options-based CCS investment evaluation model: Case study of China’s power generation sector. Applied Energy, 88(12), 4320–4333. Ziegler, D., Schmitz, K., & Weber, C. (2012). Optimal electricity generation portfolios. Computational Management Science, 9(3), 381–399. Zon, A. V., & Fuss, S. (2006). Irreversible investment under uncertainty in electricity generation: A clay–clay-vintage portfolio approach with an application to climate change policy in the UK (No. 035). United Nations University-Maastricht Economic and Social Research Institute on Innovation and Technology (MERIT).