From technology pathways to policy roadmaps to enabling measures – A multi-model approach
Introduction
The recent focus on long-term global greenhouse gas emission (GHG) mitigation has led to the production of a wide array of energy and emission specific models with varying levels of sectoral and geographic focus. On the one hand, optimisation models are beneficial in determining a technology pathway, adept at depicting what technological changes are needed in an energy system subject to a constraint, usually GHG emissions, although with little or no indication of the required policy measures, e.g., the European Commission's ‘Energy Roadmap to 2050’ [1] and the International Energy Agency's (IEA) ‘Energy Technology Perspectives’ (ETP) [2]. On the other hand, simulation models can effectively determine a policy roadmap which describe the policy steps and interim targets for emissions mitigation, although not necessarily with a focus on optimising around a certain scenario, e.g., the IEA's World Energy Outlook (WEO) [3] and the Irish ‘National Renewable Energy Action Plan’ (NREAP) [4]. Finally, analysis of these policy roadmaps can subsequently identify how enabling measures can achieve particular emission mitigation targets at a national or sectoral level through ex-ante and ex-post analysis of policies, e.g., regulations placed on car manufacturers, eco-labelling of appliances, etc. [5]. This paper brings together these three aspects in a coherent consistent iterative framework and explores the interactions, the development from one to another and highlights the need for more analysis on the effectiveness, certainty, and timing of specific measures.
The European Union (EU) face challenges in meeting emissions reduction targets in the short term (to 2020) and establishing realistic targets in the longer term (from 2030 to 2050). The European Commission's report on moving to a competitive low carbon economy in 2050 predicts that transport will be the most difficult carbon dioxide (CO2) emitting sector to decarbonise in the long-term, and is the only sector foreseen to have an increase in emissions in the medium-term [6]. Efficiency measures and biofuel blending are seen as means of meeting short-term targets (although the latter is limited by blend walls in internal combustion engines (ICE)); however, the primary challenge of decarbonising transport lies in shifting away from petroleum based liquid fuels. There is a clear and urgent need for useful methods to effectively plan and inform the implementation of policy measures to go beyond European short term targets and address this challenging long-term decarbonisation of the transport sector.
It has become common practice to address this need for planning through the integration of energy models. This integration provides results of greater value by combatting the weaknesses in one model with the strengths of another. This multi-model approach has been adopted and applied to a number of model types using varying degrees of integration. In its lightest form, two models are run independent of each other with the results of each compared until a convergence is reached giving way to a stronger result set through a low level of model structuring and a more versatile procedure than a fully integrated model, yet one that is more susceptible to errors arising due to potential inconsistencies between both model types. In the heaviest form, a complete integration of two or more models is carried out, requiring both models to be built within the same mathematical format, combatting the inconsistencies between modelling techniques, yet increasing complexity and processing power. An intermediate form creates a scaled-down representation of the structure of one model in another through integrating a reduced level of detail between model types.
A very common application of this intermediate model integration has been between computable general equilibrium (CGE) models and energy supply models, e.g., the macroeconomic model (MACRO) with a detailed energy supply model (MESSAGE) [7], and a CGE model (GEM-E3) with an energy optimisation model (TIMES) [8]. Integration of sectoral specific models have also been evident, e.g., a power systems model (PLEXOS) linked with an energy systems model (TIMES) [9], and a three-way integration of MESSAGE, TIMES, and a unit commitment optimisation tool (REMix-CEM-B) to analyse the potential of concentrated solar power in Brazil [10]. A broader, long-term analysis of the EU2030 goals was carried out with a similar analysis for Serbia combining the generic optimisation program (GenOpt) and the simulation model (EnergyPLAN) [11].
There have been very few studies dealing with the integration of transport focused models and broader energy systems models while within the few reviewed, the authors' found no representation of the individual policies necessary to achieve the policy roadmaps identified. For example, a MARKAL model of household and industry transport activities was integrated with a CGE model and outlined the potential carbon mitigation under a Kyoto target, yet gave no indication of the specific measures required [12]. A South Africa based study soft-linked five models to create long-term projections of the transport sector which consisted of developing and linking a CGE model, a vehicle parc model, a time-budget model, a freight demand model, and a fuel demand model. While this study considers the CO2 mitigation from policy roadmaps (such as shifting from private to public transport), it fails to consider the individual policies measures which may enable this shift [13].
The method of model integration presents a concise improvement from individual modelling detail and results, yet there is still a disconnect between modelling and policy analysis as described in this literature review above, especially in the area of transport, which is remarkable given the sizeable task of decarbonising transport necessary to adhere to a low carbon future. This paper aims to bridge this gap in energy modelling through (i) employing a soft-linking methodology between a least-cost optimisation model of the Irish energy system (Irish TIMES (The Irish Integrated MARKAL-EFOM System) [14]) and a sectoral simulation model of the private transport sector in Ireland (the CarSTOCK model [15]) and (ii) through using ex-post and ex-ante analysis to determine the specific enabling policy measures. Optimisation models are capable of exploring the implications of different levels of emissions reduction ambition for energy system evolution and can outline potential technology pathways; simulation models can show how particular policies and interim targets can deliver a particular energy system and hence point to policy roadmaps; finally, ex-post and ex-ante analysis facilitate analysis of enabling policy measures. The integration of these modelling and analytical approaches allows for a comprehensive description of how to decarbonise a particular sector, in this case the private car sector in the Irish energy system. The reason Ireland is chosen as a case study is twofold: first, it has the 4th highest transport emissions per capita of all EU member states (in 2014 Ireland was 2.43 tCO2/capita whereas EU average was 1.62 tCO2/capita) highlighting the onerous task of decarbonisation [16]; second, it has been a case-study for multi-modelling approach in the past, integrating Irish TIMES with the power sector [9] and the transportation sector [17].
This paper explores an ambitious long term scenario based on the European Commission's recommended CO2 greenhouse gas emissions reduction by 2050 of 80%–95% relative to 1990 [18]. This is in keeping with the Irish national policy position on climate change which declares a long-term vision guided by “an aggregate reduction in carbon dioxide (C02) of at least 80% (compared to 1990 levels) by 2050 across the electricity generation, built environment and transport sectors …” [19]. A constraint of 80% CO2 emissions reduction by 2050 relative to 1990 is entered into Irish TIMES, which determines the least-cost solution in all sectors of the economy (agriculture, residential, commercial, industry and transport). This analysis forms the basis for scenario and policy development in the CarSTOCK model, which in turn is used to analyse the type and timing of specific policy measures that can help achieve long-term decarbonisation. The efficacy of enabling policy measures requires individual scrutiny that depends on a multitude of factors which are discussed in this study – who is targeted by the measures, what type of instrument is employed, what is the timeline of these measures, and what level of change will be required. The paper is organised as follows, section 2 describes the modelling and analytical methodology, section 3 presents the results, and section 4 concludes.
Section snippets
Methods
This section first describes and defines technology pathways, policy roadmaps and enabling measures; it then describes the three technical tools employed, namely the Irish TIMES energy systems optimisation model, the CarSTOCK simulation model and ex-post analysis of policy measures; lastly, it describes the multi-model approach that integrates these three tools together.
Results
The results of the approach outlined above are presented in the following three sections: Technology Pathways – the initial results from the TIMES optimisation model, detailing the optimal technology mix within the transport sector in contributing toward an 80% reduction in CO2 emissions by 2050 relative to 1990; Policy Roadmaps – the results from the CarSTOCK model, detailing the specific policy packages necessary to contribute toward achieving the technology mix outlined by the TIMES model;
Conclusion
The soft-linking methodology employed in this study goes beyond the traditional multi-model approach by combining the foresight and comprehension of the energy system found in a least-cost optimisation model with the detailed technological representation found in sectoral simulation model with ex-post and ex-ante analysis of individual policy measures to enable long-term low-carbon solutions for the sector in question; in essence, the paper develops and aligns technology pathways to policy
Acknowledgments
This work was supported by the Environmental Protection Agency (EPA 2014-CCRP-MS.24), Innovationsfonden, Denmark (COMETS 4106-00033A) and Science Foundation Ireland (SFI) MaREI Centre (12/RC/2302). Thanks to Alessandro Chiodi for assisting with model runs, and to two anonymous reviewers and participants at the eceee summer school in 2015 at which an earlier version of this paper was presented.
References (45)
- et al.
MESSAGE–MACRO: linking an energy supply model with a macroeconomic module and solving it iteratively
Energy
(2000) - et al.
Integrated technological-economic modeling platform for energy and climate policy analysis
Energy
(2014) - et al.
Soft-linking of a power systems model to an energy systems model
Energy
(2012) - et al.
Modelling concentrated solar power (CSP) in the Brazilian energy system: a soft-linked model coupling approach
Energy
(2016) - et al.
Technology detail in a multisector CGE model: transport under climate policy
Energy Econ
(2005) - et al.
Modelling the impacts of challenging 2050 European climate mitigation targets on Ireland's energy system
Energy Policy
(2013) - et al.
Modelling private car energy demand using a stock model
Transp Res Part D Transp Environ
(March 2011) - et al.
Measuring the duration of formative phases for energy technologies
Environ Innovation Soc Transitions
(2016) - et al.
Future energy and emissions policy scenarios in Ireland for private car transport
Energy Policy
(2012) How to enhance the future use of energy policy simulation models through ex post validation
Energy
(2017)
Policy mixes for sustainability transitions: an extended concept and framework for analysis
Res Policy
Policy packaging or policy Patching? the development of complex energy efficiency policy mixes
Energy Res Soc Sci
Energy roadmap 2050
Energy technology Perspectives
World energy Outlook
Department of communications, climate action & environment
Impacts of an emission based private car taxation policy – first year ex-post analysis
Transp Res Part A
A roadmap for moving to a competitive low carbon econmoy in 2050
Simulation-based optimization of sustainable national energy systems
Energy
Quantifying the energy needs of the transport sector for South Africa: a bottom-up model
EEA greenhouse gas - data viewer
An integrated energy systems and stock modelling approach for modelling future private car energy demand
Cited by (25)
The effect of time resolution on energy system simulation in case of intermittent energies
2024, Renewable and Sustainable Energy ReviewsPolicy simulation modelling to inform national carbon budget pathways
2023, Energy Strategy ReviewsShaping the future of sustainable energy through AI-enabled circular economy policies
2023, Circular EconomyModelling the interaction between the energy system and road freight in Norway
2023, Transportation Research Part D: Transport and EnvironmentCitation Excerpt :In the conclusions, the authors underlined the possibility to couple optimisation models with agent-based models, to include modal shifts in the analysis. Also, an Irish study developed a multi-model approach using an energy systems optimisation model together with a car fleet simulation model to identify technology pathways, policy roadmaps and specific policy measures (Mulholland et al., 2017). A quantitative and systematic review of the road freight decarbonisation literature done by Meyer (2020) revealed a fast-growing and interdisciplinary research field.
Review and validation of EnergyPLAN
2022, Renewable and Sustainable Energy ReviewsBehavior matters: A systematic review of representing consumer mobility choices in energy models
2022, Energy Research and Social ScienceCitation Excerpt :The models exchanged data in both directions to utilize their combined strengths, i.e., the interaction with other sectors in the ESM and the broad choice palette for vehicle technologies in the vehicle choice model. Building upon such developments, Mulholland et al. [74,75] enrich this model coupling with the additional CIMS market share algorithm [90], representing heterogeneity and varying consumer preferences for new technologies. Tattini et al. [72] went one step further by combining models that represent behavioral realism towards car choice and modal choice.