Longterm energy models: Principles ... - Wiley Online Library

Advanced Review

Long-term energy models: Principles, characteristics, focus, and limitations ´ Gallachoir ´ ∗ Maurizio Gargiulo and Brian O There are many long-term energy models currently in use with different underlying principles, characteristics, inputs, and outputs. Over the past 30 years, considerable efforts have been made to develop new models, following different approaches that vary in terms of model starting point and on the type of questions they are designed to answer. These models focus on the period to 2050 and to 2100 and are used to build future energy scenarios to assess the impacts of policy decisions and to build a rich knowledge base for climate change and energy security policies. PET, TIAM, and MESSAGE (examples of energy system models) provide a range of technology detailed energy system configurations to deliver future energy service demands at least cost. By contrast, GEM-E3, GEMINI-E3 and, GTEM (examples of general equilibrium models) are also optimization models, but here the energy system is less detailed but the whole economy is modeled. Other models (e.g., POLES) simulate the future evolution of energy demand and supply and a number of models have been developed that include energy, but which focus on integrated assessment of the economy energy and climate response (e.g., FUND, GCAM, and MERGE). It is important for energy analysts to assess what type of model best suits the requirement and to recognize the limitations of the various models available. Significant work is required to improve linkages between models to harness respective strengths and major modeling gaps (agriculture, land C 2013 John Wiley & Sons, use, and behavior) require a specific focus for future work. Ltd.

How to cite this article:

WIREs Energy Environ 2013, 2: 158–177 doi: 10.1002/wene.62

INTRODUCTION

T

here are many ways in which long-term future energy demand and supply is modeled and a range of models and model types are currently in use. Most energy models can be classified as bottom-up technoeconomic models (which incorporate detailed engineering relationships between technology activity and energy use) or as top-down macroeconomic models.1–3 In top-down models, energy is generally modeled at an aggregate sectoral level as a derived demand that varies according as economic output and energy prices vary via elasticities. Bottom-up models represent energy sectors and technology choice

∗ Correspondence

to: [email protected]

Environmental Research Institute, University College Cork, Cork, Ireland DOI: 10.1002/wene.62

158

C

in detail, describing current and future technological options. They are generally written as mathematical programming problems therefore useful for analysis of specific changes in technology and standards. These models are not readily able to take account of price changes or macroeconomic effects. There is also a third category, hybrid models, in which bottom-up and top-down models are combined via a soft4 -link approach or a hard5–7 -link approach. The hard-link approach completely integrates two different models into a new single model. In the soft-link approach, the two models are solved in isolation and information is exchanged between them through an iterative process. Energy models may also be classified as simulation models (simulating how future energy demand and supply trends will evolve based on projected trends of energy drivers) or optimization models (in which, e.g., future energy demand is delivered as least

2013 John Wiley & Sons, Ltd.

Volume 2, March/April 2013

WIREs Energy and Environment

Long-term energy models: Principles, characteristics, focus, and limitations

cost) depending on the model formulation. Further classification is possible for optimization models depending on whether the energy system is optimized (partial equilibrium, also known as energy system models, which are technology detailed) or whether the optimization takes place economy wide (general equilibrium models, in which the energy system is more simply described). A significant focus of current long-term energy modeling activity is to inform climate policy. Energy is also a key component of climate modeling, given the significant role of energy in terms of both the contribution to current climate change and to future mitigation options. Given this interplay between energy and climate modeling, this review on long-term energy models focuses on energy system models and on the energy dimension of climate models. The paper provides a review of models that currently exist rather than a review of energy modeling techniques. One important branch of models considered in this review is energy system models. Approaching energy as a system instead of a set of elements gives the advantage of identifying the most important substitution options that are linked to the system as a whole and cannot be understood looking at a single technology or commodity or sector. For example, a focus on the electricity sector risks excluding possible unforeseen step changes in electricity demand, for example, due to the electrification of transport or of heating. By considering energy supply and demand across all sectors simultaneously, systems analysis applies systems principles to aid decision makers in problems of identifying, quantifying, and controlling a system. Although taking into account multiple objectives, constraints, resources, it aims to specify possible courses of action, together with their risks, costs, and benefits. Current energy systems are the result of complex country dependent, multisector developments. Over the past 30 years, considerable efforts have been made to develop new models8–27 for individual sectors22–27 or integrated models for more than one sector,10–21 following different approaches that depend on the starting point for the model (e.g., economics, technology, or integrated assessment) and on the type of questions they are designed to answer. These models have become standard tools for scenario and policy analysis both for industrialized and developing countries (with different aims). The key policy focus has shifted from energy security to climate-change mitigation but more recently energy security has reemerged as a focus topic. This paper provides a state-of-the-art review of the currently used energy models, focusing on essential principles, outlining their potential, and acknowl-


C

edging fundamental limitations, for the interested interdisciplinary readership. The models that are currently used are taken as the starting point, rather than the model types previously outlined, although some classification is applied in the model comparison after the models are individually described. A number of the models described in this paper are used for climate modeling within the Energy Modelling Forum,28–30 some were developed through collaboration under the auspices of the International Energy Agency and others have been developed and used in research projects focusing on the European Union (EU).

CURRENT ENERGY MODELS This section provides an overview of the main energy models currently in use. Table 1 provides a useful summary of these models, including key attributes for each model that includes geographic coverage, typical time horizon, model type, focus, and some applications. The order in which the models are discussed here follows that of Table 1 with the initial focus on bottom-up energy models, then top-down models for climate-change analysis, followed by integrated assessment models for a global focus for long and medium term analysis and finally a smaller group of models with different approaches.

MESSAGE Model for Energy Supply Strategy Alternatives and their General Environmental Impact31 (MESSAGE) is a dynamic linear programming model, calculating cost-minimal supply structures under the constraints of resource availability, the menu of given technologies, and the demand for useful energy. MESSAGE is a systems engineering optimization model used for medium to long-term energy system planning, energy policy analysis, and scenario development.32–34 The model provides a framework for representing an energy system with all its interdependencies from resource extraction, imports and exports, conversion, transport, and distribution, to the provision of energy end-use services such as light, space conditioning, industrial production processes, and transportation across 11 macroregions (North America, Western Europe, Pacific OECD, Central and Easter Europe, Newly Independent States of the former Soviet union, Centrally Planned Asia and China, South Asia, Other Pacific Asia, Middle East and North Africa, Latin America and the Caribbean and Sub-Saharan Africa). For a detailed description of each macroregion, the reader is referred to the MESSAGE website.35


159

160

World

Europe extended

Europe extended

MESSAGE

PET

PRIMES

C


World

World

GTAP / GTAP-E

World

ETSAP-TIAM/ TIAM-WORLD

POLES

World

WITCH

CIMS

Geographic Focus

Model Name

T A B L E 1 Energy Models List

113

18

15 / 16

12

EU27, Norway, Switzerland, South east Europe Canada, USA, and China

EU27, Norway, Switzerland, Iceland and Six Balkan Countries

11

Number of Regions/ Macroregions

2004 (or 2007)

2030

2005–2100

2005–2030

2000–2050

2005–2050

1990–2060

Time Horizon

Top-down/CGE

Top-down/ Econometric

Bottom-up/ OptimizationIntegrated Assessment Model

Bottom-up/Top-down/ Integrated Assessment Model

Bottom-up/Top-down

Bottom-up/ Top-down

Bottom-up/ Optimization

Bottom-up/ Optimization

Type of Model

Climate-change policies

Energy modeling, energy policy analysis, environmental targets and scenario analysis

Energy modeling, energy policy analysis, environmental targets and climate-change analysis Energy modeling, energy policy analysis, environmental targets and climate-change analysis

Energy modeling, energy policy analysis, environmental targets and scenario analysis Energy modeling, energy policy analysis

Energy modeling, energy policy analysis, environmental targets and scenario analysis Energy modeling, energy policy analysis, environmental targets and scenario analysis

Focus

(Continued)

In assessment, modeling activities and for climate-change analysis. The model has been used also for Regional economic and energy implications of reaching global climate targets – A policy scenario analysis To support the World Energy Technology 2030 report, the WETO-H2 2050 report and the quantitative scenarios of the World Energy Council in 2007 To assess environmental and energy issues

To evaluate the effectiveness and economic impact of public policies to reduce greenhouse gas emissions In assessment and modeling activities for climate-change mitigation analysis

To develop energy technology strategies for carbon dioxide mitigation and sustainable development To evaluate energy scenarios, environmental and renewable targets in EU Projects (NEEDS, RES2020, REACCESS, REALISEGRID, COMET, Irish-TIMES) To evaluate the set of policies and measure for the European Member states

Model Applications

Advanced Review wires.wiley.com/wene



C

World

World

World Europe

MERGE

LEAP

World

GTEM

FUND

World and Europe

GEM-E3

World

World

GEMINI-E3

GCAM (formerly MiniCAM)

Geographic Focus

Model Name

T A B L E 1 Continued


9

16

14

13

21 W/24 E

28

Number of Regions/ Macroregions

2000–2150

1950–3000

1990–2095

1997–2100

2025/2050

Time Horizon

Accounting framework

Integrated Assessment model

Integrated Assessment Model

Integrated assessment model

Top-down general equilibrium model

Top-down/CGE

Top-down/CGE

Type of Model

Energy modeling, energy policy analysis, environmental targets and scenario analysis

Impacts of climate change and to perform cost-benefit and cost-effectiveness analyses of greenhouse gas emission reduction policies Climate-change policies

Energy modeling, land use, energy policy analysis, environmental targets and scenario analysis




Focus

(Continued)

To evaluate regional and global effects of GHG reduction policies To evaluate energy scenarios, environmental and renewable targets

To assess European and world climate-change policies at the microeconomic and the macroeconomic levels To assess European and world climate-change policies at the microeconomic and the macroeconomic levels To evaluate the economic impact of climate-change policy: the role of technology and economic instruments In assessment and modeling activities such as the Energy Modeling Forum (EMF), the U.S. Climate Change Technology Program, and the U.S. Climate Change Science Program and IPCC assessment reports To advice policymakers about proper and not-so-proper strategies

Model Applications

WIREs Energy and Environment Long-term energy models: Principles, characteristics, focus, and limitations

161

To produce the EIA International Energy Outlook 2007 To explore the long-term dynamics of global change as the result of interacting demographic, technological, economic, social, cultural and political factors To assess climate-change policy analysis

Minimization of the total system costs is the default objective function in MESSAGE. Scenarios are developed by MESSAGE through minimizing the total systems costs under the constraints imposed on the energy system. Given this information and other scenario features such as the demand for energy services, the model configures the evolution of the energy system from the base year to the end of the time horizon (in 10 year steps). It provides the installed capacities of technologies, energy outputs and inputs, energy requirements at various stages of the energy systems, costs, and emissions. The degree of technological detail in the representation of an energy system is flexible and depends on the geographical and temporal scope of the problem being analyzed. A typical model application is constructed by specifying performance characteristics of a set of technologies and defining a Reference Energy System (RES) http://www.iiasa.ac.at/ Research/ECS/images/res1a.jpg to be included in a given study/analysis that includes all the possible energy chains that the model can make use of. In the course of a model run, MESSAGE then determines how much of the available technologies and resources are actually used to satisfy a particular end-use demand, subject to various constraints, whereas minimizing total discounted energy system costs. The MESSAGE model is resident at the International Institute for Applied Systems Analysis (IIASA) in Austria.

Dynamic recursive model 2005–2100

PET Model The Pan European Times (PET) Model36 is a multiregional partial equilibrium model of Europe built with TIMES,37 the technical economic model of IEAETSAP.38 The PET model represents the energy system of 36 European regions (EU27, Iceland, Norway, Switzerland, and six Balkan countries) and its possible long-term evolution and was developed following a series of EC funded projects (NEEDS,39 RES2020,40 REACCESS,41 REALISEGRID,42 COMET, IrishTIMES).43 PET is set up to explore the development of the European energy system till 2050. However, it is easily extended into the future, provided that the necessary data can be provided. The current version of the model was recalibrated to 2005 Eurostat data. Annual flows of electricity are split by four seasons and daily load profiles (night, day, and peak). The seasonal dependency is extended to heat and partly natural gas. Each technology can operate in the same twelve time-slices (fractions of a year) describing the four seasons, and day/night/peak times. At present,

World Phoenix (formerly SGM)

24

Simulation World IMAGE

24

1970–2050 (2100)

Energy modeling, energy policy analysis Climate-change analysis Econometric model 2030 World WEPS+

16

Type of Model Number of Regions/ Macroregions Geographic Focus Model Name

T A B L E 1 Continued 162

wires.wiley.com/wene

Time Horizon

Focus

Model Applications

Advanced Review

C





processes producing and consuming electricity have the more detailed level (12 time slices), whereas the processes involving other commodities are set for seasonal operation. Each region is described and modeled in its supply sector (fuel mining, primary, and secondary production, exogenous import and export), its power generation sector (including also the combined heat and power description and the heat production by district heating plants), and its demand sectors (residential, commercial, agricultural, transport, industrial). The model includes CO2 and non-CO2 emissions. Furthermore, emissions are broken down by emitting sector (energy use in agriculture, commercial, power sector, industry, residential, and transport). The trade of energy between the regions of PET is endogenously modeled for: electricity, natural gas, biomass, and CO2 emissions. The model represents two complementary sets of system elements: technical aspects, which include energy, emissions and engineering, and economic aspects. The representation assumes that properties of both aspects hold. The PET model uses the partial equilibrium version of TIMES, where the demand for energy services depends endogenously on own price elasticities. In other words, it is assumed that the system develops maintaining intratemporal and intertemporal partial economic equilibrium and always occupies the technical possibility frontier. There are different versions of the models resident in different places. The PEM model (30 regions) is resident at the University of Stuttgart, a 30 region version of PET is resident at the Katholieke Universiteit Leuven and at the Joint Research Center in Seville, whereas the PET 36 region versions is resident at Kanlo (France), Kanors (India), and E4SMA (Italy).

PRIMES Model The PRIMES model is a partial equilibrium model simulating the entire energy system, both in demand and supply. It is a mixed bottom-up and top-down model based on separate modules for each demand and supply sector and separate decision making. The first version of the model has been supported by a series of research programs of the European Commission. The PRIMES model represents the EU 27 Member states and candidate member states and neighbors (Norway, Switzerland, Turkey, and South East Europe). The model has been used in the evaluation of the set of policies and measure for the European Member states.44


C

PRIMES is a modeling system that simulates a market equilibrium solution for energy supply and demand in the EU member states.45, 46 The model determines the equilibrium by finding the prices of each energy form such that the quantity producers find best to supply match the quantity consumers wish to use. The equilibrium is static (within each time period) but repeated in a time-forward path, under dynamic relationships. The model is behavioral but also represent in an explicit and detailed way the available energy demand and supply technologies and pollution abatement technologies. The system reflects considerations about market economics, industry structure, energy/environmental policies, and regulation. It covers a medium to long-term horizon PRIMES is organized around a modular design representing in a different manner fuel supply, energy conversion, and end-use of demand sectors. The individual modules vary in the depth of their structural representation. For example, the electricity module covers the whole Europe, while representing chronological load curves and dispatching at the national level. The natural gas market also expands over the whole Europe. However, coal supply, refineries, and demand operate at the national level. Furthermore, the modularity allows any single sector or group of sectors to be run independently for stand-alone analysis. In PRIMES, producers and consumers both respond to changes in price. The factors determining the demand for and the supply of each fuel are analyzed and represented, so they form the demand and/or supply behavior of the agents. Through an iterative process, the model determines the economic equilibrium for each fuel market. Price-driven equilibrium is considered in all energy and environment markets, including Europe-wide clearing of oil and gas markets, as well as Europe-wide networks, such as the Europewide power grid and natural gas network. The PRIMES model is resident at E3MLAB National Technical University of Athens (Greece).

CIMS CIMS is an energy-economy simulation model developed and maintained by the Energy and Materials Research Group at Simon Fraser University in Canada. The CIMS model can be used to evaluate the effectiveness and economic impact of public policies to reduce greenhouse gas (GHG) emissions in Canada, USA, and China.47, 48 The model is used to simulate various policy scenarios and to understand how climate policy affects energy trade. CIMS is a simulation model that integrates bottom-up technology modeling


163

Advanced Review


and top-down economic modeling capabilities and so has a detailed technology database (1500 technologies) for the 15 economic sectors. The time horizon of the model is 2000–2030 in which the model simulates the evolution of technologies in 5-year steps. The scenarios analysis developed with the CIMS model can provide information on the cost, effectiveness, and economic impacts of climate policies. CIMS simulates the response of energy demand to changes in prices in its sectors, forecasting energy demand and emissions based on technologies choice that are consuming different energy carriers. It uses a market share function to simulate real-world preferences and realistic decision-making behavior. The model can simulate environmental policies and energy trade (limited by assumptions) playing with technology costs. The actual model simulates, based on demand change and energy production cost, the trade of crude oil, natural gas, electricity, and refined petroleum products between Canada, the USA, and the rest of the World. The technology substitution in the model is simulated with 5-year steps in which the model from one period to the next change a stock portion based on the retirement profile and the new available technologies. The share of each new technology is determined by its lifecycle cost (investment, operating, and maintenance costs). The CIMS model uses a market share function to determine how much of each technology is adopted. This is a way of simulating real-world preferences and decision-making behavior, which is heterogeneous and may be based on multiple factors and constraints other than costs. The CIMS model requires multiple assumptions about energy supply and energy markets to calculate changes in the costs of technologies on the supply side (production, conversion, and transportation) and this is the limitation of this model to be applied to larger regions and trade between more regions.

WITCH The World Induced Technical Change Hybrid49–51 (WITCH) is a regionally disaggregated (12 macro regions) hard-link hybrid global model with a top-down optimal growth structure and a detailed energy input component (bottom-up) for the electricity sector. It is an inter-temporal (solved for 30 5-year periods) general equilibrium computation model, but it is enriched by the presence of key technologies in the energy sector, as well as by the modeling of endogenous technological learning and innovation. An integrated climate module makes it possible to track changes

164

C

in atmospheric CO2 concentrations and world mean temperatures as a consequence of the use of fossil fuels and feeds a damage function, which in turn delivers the effect of climate changes on the economy. The model features the main economic and environmental policies in each world region as the outcome of a dynamic game. WITCH belongs to the class of Integrated Assessment Models as it possesses a climate module that feeds climate changes back into the economy. In WITCH, policy decisions adopted in one region of the world affect what goes on in all the other regions. This implies that the equilibrium of the model, that is, the optimal inter-temporal investment profiles, R&D strategies, and direct consumption of natural resources, must be computed by solving a dynamic game. WITCH is designed to analyze optimal climate mitigation strategies within a gametheoretical framework, while portraying the evolution of energy technologies with adequate detail and allowing for endogenous technological progress. It is a ‘hard-link hybrid’ model in the sense that the energy sector is contained within the economy: capital and resources for energy generation are therefore allocated optimally with respect to the whole economy. Optimization growth models are usually very limited in terms of technological detail. This severely constrains the analysis of climate-change issues, which are closely related to the evolution of energy sector technologies. In WITCH, this component is considerably richer in information than in most macrogrowth models, although still much simpler than that of large scale energy system models. WITCH is resident at Fondazione ENI Enrico Mattei (FEEM, Italy).

ETSAP-TIAM and TIAM-World The TIMES Integrated Assessment Model (TIAM)52, 53 is a multiregional partial equilibrium model of the World built with TIMES, the technical economic modeling tool of IEA-ETSAP.38 ETSAP-TIAM and TIAM-World are two different versions of the TIAM model. The first one is resident at ETSAP and available for all the ETSAP members and the second one is resident at Kanlo Consultants Sarl ` (France) and KanORS-EMR (India). TIAM represents the energy system of the World divided in 15 regions (ETSAP-TIAM) and 16 regions (TIAM-World). The model contains explicit detailed descriptions of more than one thousand technologies and one hundred commodities in each region, logically interrelated in a RES. TIAM is driven by





a set of 42 demands for energy services in all sectors: agriculture, residential, commercial, industry, and transportation. Demands for energy services are specified by the user for the Reference scenario, and have each an own price elasticity. Each demand varies endogenously in alternate scenarios, in response to endogenous price changes. The model thus computes a dynamic intertemporal partial equilibrium on worldwide energy and emission markets based on the maximization of total surplus, defined as the sum of surplus of the suppliers and consumers. TIAM also integrates a climate module permitting the computation and modeling of globally averaged temperaturechange limits related to concentrations, radiative forcing, and temperature increase. For each region, the supply sector (fuel mining, primary and secondary production, and exogenous import and export), the power generation sector (including also the combined heat and power description and the heat production by district heating plants), and the demand sectors (residential, commercial, agricultural, transport, industrial) are modeled in a very detailed manner. Such technological detail allows precise tracking of capital turnover, and provides a precise description of technological competition. A coupled use of TIAM-WORLD and the macro-economic model GEMINI-E3 has also been developed, permitting the combination of the strengths of the two models54, 55 TIAM-World was also hardlinked with the PET model (REACCESS41 project), resulting in a global 51 region model where energy corridors were modeled in a detailed manner to assess energy security of EU.

POLES The Prospective Outlook on Long-term Energy Systems56 (POLES) model is an econometric, partialequilibrium World model (equilibrium between energy demand and supply, but economic assumptions remain exogenous). The model has been developed first by CNRS (France) and now Enerdata, in collaboration with LEPII (formerly IEPE—Institute of Energy Policy and Economics) and IPTS (Spain, European Commission Joint Research Centre), coordinates studies on long-term energy outlooks at world level with the POLES model. POLES allows the projections of energy demand and supply by region/country and international oil/gas/coal prices, the simulation of technology development for electricity supply and simulation of CO2 emissions and analysis of CO2 abatement policies and carbon values.


C

The model simulates the energy demand and supply for 32 countries and 18 world regions. There are 15 energy demand sectors (main industrial branches, transport modes, residential, and service sectors), about 40 technologies of power and hydrogen production. It works in a year-by-year recursive simulation and partial equilibrium framework, with endogenous international energy prices and lagged adjustments of supply and demand by world region. For the demand, behavioral equations take into account the combination of price and revenue effects, technoeconomic constraints and technological trends. Oil and gas supply profiles are projected for the main producing countries from a simulation of the drilling activity and discovery of new reserves, given the price, the existing resources and the cumulative production. The integration of import demand and export capacities of the different regions is included in the international energy market module, which balances the international energy flows. The changes in international prices of oil, gas, and coal are endogenous, taking into account the Gulf capacity utilization rate for oil, the reserve on production ratio for oil and gas, and the trend in productivity and production costs for coal. The can be used to produce long-term (2030) world energy outlooks with demand, supply and price projections by main region, CO2 emission marginal abatement cost curves by region, and emission trading systems analyses, under different market configurations and trading rules and technology improvement scenarios (with exogenous or endogenous technological change) and analyses of the value of technological progress in the context of CO2 abatement policies. The POLES model has been used in many forecasting studies, at both national and international levels. In particular it supported the World Energy Technology 203057 report (published in 2003) and the WETO-H2 205058 report (published in 2007) for the European Commission, as well as the quantitative scenarios of the World Energy Council59 in 2007. Target users of the model are international organizations and policy.

GTAP and GTAP-E The Global Trade Analysis Project (GTAP)60 is a research program initiated in 1992 to provide the economic research community with a global economic dataset for use in the quantitative analyses of international economic issues. The Project’s objectives


165

Advanced Review


include the provision of a documented, publicly available, global, general equilibrium data base. The GTAP research program is coordinated by Thomas Hertel, Director of the Center for Global Trade Analysis at Purdue University. The GTAP version8 database represents global production and trade for 129 regions, 57 commodities. The data characterize intermediate demand and bilateral trade in 2004 (or 2007), including tax rates on imports and exports. The GTAP Data Base is most commonly used with the GTAP Model. First, the user must aggregate the data (regions, commodities, and endowments) to the desired level and then use with the GTAP model to analyze the impact of global policies (trade, environmental, migration policies are commonly examined). Alternatively a user may be interested in extracting country SAMs or I-O tables from the GTAP Data Base for single country models. The GTAP model is a multiregion, multisector, computable general equilibrium model, with perfect competition, and constant returns to scale. Bilateral trade is handled via the Armington assumption. The GTAP Model also gives users a wide range of closure options, including unemployment, tax revenue replacement and fixed trade balance closures, and a selection of partial equilibrium closures (which facilitate comparison of results to studies based on partial equilibrium assumptions). The GTAP model includes the treatment of private household preferences using the nonhomothetic CDE functional form, the explicit treatment of international trade and transport margins. Bilateral trade is handled via the Armington assumption. A global banking sector that intermediates between global savings and consumption. The GTAP Model also gives users a wide range of closure options, including unemployment, tax revenue replacement and fixed trade balance closures, and a selection of partial equilibrium closures (which facilitate comparison of results to studies based on partial equilibrium assumptions).

GTAP-E The extended version of GTAP that incorporates energy substitution into the standard database is called GTAP-E.61 In addition, GTAP-E incorporates carbon emissions from the combustion of fossil fuels and this revised version of GTAP-E provides for a mechanism to trade these emissions internationally. The policy relevance of GTAP-E in the context of the existing debate about climate change is illustrated by some simulations of the implementation of the Kyoto Pro-

166

C

tocol. It is hoped that the proposed model will be used by individuals in the GTAP network who may not be themselves energy modelers, but who require a better representation of the energy–economy linkages than is currently offered in the standard GTAP model.

GEMINI-E3 GEMINI-E362, 63 is a Computable General Equilibrium (CGE) Model and represents now a family of models of different specifications and with several successive versions. It retains many specifications that are common to CGE models but also some specific features, mainly concerning the measurement and analysis of the welfare cost of policies and the great detail in the representation of taxation and social security contributions. The model can be adapted to specific contexts indeed CGE models, beside their main virtue that is total consistency at the domestic and at the world levels, are very flexible in their specification. GEMINI-E3 is the name of the first CGE model developed jointly by the French Ministry of Equipment and the French Atomic Energy Agency. The fifth version has been developed with the collaboration of the Swiss Federal Institute of Technology (Lausanne). The model started in 1994 at the French Energy Atomic Agency and presently is developed and managed in collaboration between ASSESSECO and EPFL (REME). GEMINI-E3 has been specifically designed to assess European and world climate-change policies,63 in particular the effects of the Kyoto Protocol,64 both at the microeconomic and the macroeconomic levels. It is a multicountry (28), multisector (18), dynamic model incorporating a highly detailed representation of indirect taxation. It has been thoroughly used since 1995 to analyze a wide range of questions linked to energy, environment, and economic growth. The model runs in annual steps mainly from the base year 2001 to 2025, although some longer term scenarios—up to 2050—have also been simulated. The present version of the model is capable of assessing intra-European and domestic policies such as the directive on quotas, project of directive ‘energyclimate’ and the determination of carbon value. A new version, GEMINI-EMU, has been developed specifically aimed at assessing intra-European macroeconomic policies but also relevant for climate-change scenarios appraisal. As most CGE models, GEMINI-E3 simulates all relevant markets, domestic and international, considered as perfectly competitive, which implies that the corresponding prices are flexible. Time periods are





linked in the model through endogenous real rates of interest determined through the balancing of savings and investment. Goods of the same sector produced by the different countries are not supposed to be perfectly competitive. The main outputs of the GEMINI-E3 model are by country and annually: carbon taxes, marginal abatement costs and prices of tradable permits (when relevant), effective abatement of CO2 emissions, net sales of tradable permits (when relevant), total net welfare loss and components (net loss from terms of trade, pure deadweight loss of taxation, net purchases of tradable permits when relevant), macroeconomic aggregates (e.g., production, imports and final demand), real exchange rates and real interest rates, and data at the industry level (e.g., change in production and in factors of production, prices of goods).

GEM-E3 The General Equilibrium Model for Energy– Economy–Environment interactions65–69 (GEM-E3) is a computable general equilibrium model. The model has been developed as a multinational collaboration project, partly funded by the European Communities, DG Research, 5th Framework program and by national authorities, and further developments are continuously under way. Applications of the model have been (or are currently being) carried out for several Directorate Generals of the European Commission (economic affairs, competition, environment, taxation, research) and for national authorities. There are two versions of GEM-E3: GEM-E3 Europe and GEM-E3 World. They differ in their geographical and sectoral coverage, but the model specification is the same.70 The world version of GEM-E3 is based on the GTAP7 database (base year 2004). The regional aggregation is flexible and allows for the individual representation of all major world economies. The European version covers 24 EU countries (all EU Member States, except for Luxemburg, Malta, and Cyprus) and the rest of the world (in a reduced form) and it is based on EUROSTAT data. GEM-E3 aims at covering the interactions between the economy, the energy system, and the environment. The model computes simultaneously the competitive market equilibrium under the Walras’ law and determines the optimum balance for energy demand/supply and emission/abatement. The structural features of the energy/environment system and the policy-oriented instruments (e.g., taxation) have a considerable level of detail. The geographical regions are linked through bilateral trade. The prices are


C

computed by the model as a result of supply and demand interactions in the markets and different market clearing mechanisms, in addition to perfect competition, are allowed. The model is simultaneously multinational (for the EU or the World) and specific for each country/region; markets clear European/World-wide, while country/region-specific policies and distributional analyses are supported. The model is disaggregated concerning sectors, structural features of energy/environment, and policy-oriented instruments (e.g., taxation). The model formulates production technologies in an endogenous manner allowing for price-driven derivation of all intermediate consumption and the services from capital and labor. GEM-E3 is dynamic, recursive over time, driven by accumulation of capital and equipment. Technology progress is explicitly represented in the production function, either exogenous or endogenous, depending on R&D expenditure by private and public sector and taking into account spill over effects. The model formulates pollution permits for atmospheric pollutants and flexibility instruments allowing for a variety options, including: allocation (grandfathering, auctioneering, etc.), user-defined bubbles for traders, various systems of exemptions, various systems for revenue recycling, and so on. The model setup includes the energy-related and nonenergy related emissions of carbon dioxide (CO2 ) and other GHG.

GTEM The Global Trade and Environment Model (GTEM)71 is a top-down, dynamic, multiregion, multisector, general equilibrium model of the world economy. The model was developed by ABARES specifically to address policy issues with long-term global dimensions. In the past, GTEM has been used to analyze issues such as the climate-change response policies including the Kyoto Protocol, trade reform under the World Trade Organization, and trends and issues in international commodity and energy markets. GTEM represents the global economy through 13 regions (Australia, US, China, India, EU 25, Japan, Indonesia, Other South and East Asia, Former Soviet Union, OPEC, Canada, South Africa, and rest of the World) each with 19 industrial sectors and a representative household (for each whole regional society). Trade and investment link the regions and a range of taxes and subsidies capture government policies. The model assumes multiple production technologies for three energy-intensive sectors: the electricity, transport, and iron and steel sectors. For a detailed description of each macroregion, the reader is referred


167

Advanced Review


to the host website.72 The model horizon is 1997– 2100 in three 1-year steps Climate-change analysis and global energy analysis can be addressed with GTEM. Climate-change analysis using GTEM can include the modeling of market based policy instruments either in the form of ‘cap and trade’ schemes (such as, the Kyoto Protocol) or carbon taxes. The modeling of a domestic and international emissions trading can feature options for the banking of emissions quota and market power, carbon sinks, or some well-known flexible mechanisms (e.g., the clean development mechanism in the case of Kyoto Protocol). GTEM can also be used to produce emission projections for various sectors and examine the impacts of climate-change regulatory strategies, such as mandatory renewable energy targets. Global energy analysis using GTEM can includes the impacts of policies likely to affect global energy markets, such as energy subsidies and carbon taxes. GTEM can be used for analyzing complex trade policy issues because it takes into account the interactions between different sectors in each economy and between economies through bilateral trade flows in goods and services. Also, GTEM models the interactions among policies, namely domestic support, export subsidies and import tariffs that significantly affect international trade. The model provides estimates of the impacts of policy changes on key economic variables including trade and investment flows, the price of consumer goods and inputs to production, sectoral and regional output and, ultimately, regional income and expenditure. ABARES has used GTEM extensively for analysis of agricultural trade liberalization issues in the context of the WTO negotiations.73, 74

GCAM (Formerly MiniCam) The Global Change Assessment Model75 (GCAM, formerly MiniCAM) is a partial-equilibrium model (energy and land-use) including numerous energy supply technologies, agriculture and land use model, and a reduced form climate model. Emissions include CO2 , CH4 , N2 O, and SO2 . The time horizon of the model is 1990–2095 with variable time steps. GCAM includes three end-use sectors (buildings, industry, and transportation). Energy supply and transformation sectors: fossil-fuels (oil, natural gas, and coal), biomass (traditional and modern), electricity, hydrogen, and synthetic fuels. The model covers 14 regions (US, Canada, Western Europe, Japan, Australia & New Zealand, Former Soviet Union, Eastern Europe, Latin America, Africa, Middle East,

168

C

China, India, South Korea, and Rest of South & East Asia). GCAM integrates four existing models, the Edmonds, Reilly, Barns (ERB)76 energy economic model and the agriculture, forestry, and land use model (ALM)77 to represents long-term trends in economic output, energy use, and GHG emissions through detailed submodules representing energy resources, primary energy supply and demand, energy markets including world trade and electricity conversion. The atmospheric composition and global climate changes is evaluated using the HECTOR78 model, and SCENGEN79 , which models the regional patterns of climate change. GCAM enhances the ability to understand the impact of technologies and policies related to GHG emissions in a national and global context, including the ability to quickly evaluate technologies including carbon sequestration.80, 81 In the model, biomass land competes with food and fiber uses in the agriculture/land-use model. The HECTOR component provides GHG concentrations, radiative forcing, and climate change. The flexible object-oriented structure allows new technologies and sectors to be quickly implemented. The model produces emissions of GHGs (carbon dioxide, methane and nitrous oxide) and other radiatively important substances such as sulfur dioxide and, examines the consequences of these emissions for climate change and sea-level. The agriculture-land-use model (AgLU) endogenously determines land use, land cover, and the stocks and flows of carbon from terrestrial reservoirs. AgLU is fully integrated with the GCAM energy and economy modules. The model data for the agriculture and land use parts of the model is composed of 151 subregions in terms of land use, based on a division of the extant agroecological zones (AEZs) within each of GCAM’s 14 global geopolitical regions. The model is designed to allow specification of different options for future crop management for each crop in each subregion. Stocks and flows of terrestrial carbon and other GHGs are determined by associated land use and land cover and land-use-land-cover changes. The GCAM model is developed and resident at the Joint Global Change Research Institute at Pacific Northwest Laboratory, Maryland.

FUND The Climate Framework for Uncertainty, Negotiation and Distribution82 (FUND) is an integrated assessment model of climate change. FUND was originally set up to study the role of international capital





transfers in climate policy, but it soon evolved into a test-bed for studying impacts of climate change in a dynamic context, and it is now often used to perform cost-benefit and cost-effectiveness analyses of GHG emission reduction policies, to study equity of climate change and climate policy, and to support game-theoretic investigations into international environmental agreements. FUND is an integrated assessment model of projections of populations, economic activity and emissions, carbon cycle and climate model responses, and estimates of the monetized welfare impacts of climate change.83, 84 Climate-change impacts are monetized in 1995 dollars and are modeled over 16 regions. Modeled impacts include agriculture, forestry, sea level rise, cardiovascular, and respiratory disorders influenced by cold and heat stress, malaria, dengue fever, schistosomiasis, diarrhea, energy consumption, water resources, unmanaged ecosystems, and tropical and extratropical storm impacts. The model version 3.785 distinguishes 16 major regions of the world, viz. the United States of America (USA), Canada, Western Europe, Japan and South Korea, Australia and New Zealand, Central and Eastern Europe, the former Soviet Union, the Middle East, Central America, South America, South Asia, Southeast Asia, China, North Africa, SubSaharan Africa, and Small Island States. The model runs from 1950 to 3000 in time steps of one year. The prime reason for starting in 1950 is to initialize the climate-change impact module. In FUND, some of the impacts of climate change are assumed to depend on the impact of the previous year, this way reflecting the process of adjustment to climate change. Because the initial values to be used for the year 1950 cannot be approximated very well, both physical and monetized impacts of climate change tend to be misrepresented in the first few decades of the model runs. The centuries after the 21st are included to assess the long-term implications of climate change. Previous versions of the model stopped at 2300. FUND is used to advice policymakers about proper and not-so-proper strategies.86 The model, however, always reflects its developer’s world views. FUND was originally developed by Richard Tol. It is now codeveloped by David Anthoff and Richard Tol. FUND does not have an institutional home.

MERGE The Model for Evaluating Regional and Global Effects (MERGE)87 is an integrated assessment model that provides a framework for assessing climatechange management proposals. The world modeled


C

in MERGE is divided into nine geopolitical regions: Canada, Australia and New Zealand (CANZ); China; eastern Europe and the former Soviet Union (EEFSU); India; Japan; Mexico, and OPEC (MOPEC); Western Europe (WEUR); USA; and the rest of the world (ROW). MERGE is a model for estimating the regional and global effects of GHG reductions.88 It quantifies alternative ways of thinking about climate change. The model is sufficiently flexible to explore alternative views on a wide range of contentious issues: costs of abatement, damages of climate change, valuation, and discounting. MERGE contains submodels governing: the domestic and international economy energy-related emissions of GHGs nonenergy emissions of GHG’s global climate change—market and nonmarket damages. The MERGE model runs in 10-year intervals from 2000 through the horizon date of 2150, includes non-CO2 GHGs, sinks, and abatement technologies, extrapolating technical progress in abatement post2010. The technological database is benchmarked with US Energy Information Agency, International Energy Outlook, 2003. Price responsiveness is introduced through a top-down production function. Output depends upon the inputs of capital, labor, and energy. Energyrelated emissions are projected through a bottomup perspective. Separate technologies are defined for each source of electric and nonelectric energy. Fuel demands are estimated through ‘process analysis’. Each period’s emissions are translated into global concentrations and in turn to the impacts on mean global indicators such as temperature change. MERGE may be operated in a ‘cost-effective’ mode— supposing that international negotiations lead to a time path of emissions that satisfies a constraint on concentrations or on temperature change. The model may also be operated in a ‘benefit-cost’ mode—choosing a time path of emissions that maximizes the discounted utility of consumption, after making allowance for the disutility of climate change.

LEAP (Applications) The Long range Energy Alternatives Planning System89 (LEAP), is a widely used software tool for energy policy analysis and climate-change mitigation assessment. It is an accounting framework that can be used to build energy models and is updated by the Stockholm Environment Institute and has been used at many different scales ranging from cities and states to national, regional, and global applications.21


169

Advanced Review


LEAP models are integrated modeling tools that track energy consumption, production, and resource extraction in all sectors of an economy. These are models built for medium long-term analysis (20–50 years) with most of its calculations at annual time step. Some results are calculated with a finer level of temporal detail (e.g., in the electric sector the year is split in time slices). These models are used to account for both energy sector and nonenergy sector GHG emission sources and sinks. These models include accounting and simulations methodologies that can incorporate and use data also from other specialized models. Examples include work by the Chinese Energy Research Institute (ERI) who used LEAP to explore how China could achieve its development goals while also reducing its carbon intensity and work in the US, where the Natural Resources Defense Council (NRDC) uses LEAP to analyze national fuel economy standards and advocate for policies that encourage clean vehicles and fuels. These studies have helped to influence national energy policies and plans.90, 91 Recently, LEAP models have integrated a new electricity optimization module based on the open source modeling tool OSeMOSYS.20

in GDP and energy prices. The model starts off by forecasting the end-use consumption of several energy sources for each sector in each of the regions. The sectors are residential, commercial, industrial, and transportation. The energy sources are oil, natural gas, coal, and electricity. These are forecasted based on the coefficients earlier calculated from the reference case and the levels of GDP and energy prices. The model specification uses a real GDP elasticity, an own-price elasticity, and (where relevant) a price of oil to natural gas elasticity. For electricity demand the model specification uses only a real GDP elasticity. Separately the model calculates the supply for oil, natural gas, and coal in each of the 16 world regions. The supply is forecasted based on the coefficients earlier calculated from the reference case and the levels of each energy price. The World Industrial Model (WIM)92, 93 of the World Energy Projection System Plus (WEPS+) is an energy demand modeling system that projects industrial end-use sector energy consumption for 16 international regions and it was used to produce the industrial sector projections published in the International Energy Outlook 2010 (IEO2010). WIM is one of 13 components of WEPS+ energy modeling system, but the industrial model can also be run as a separate, individual model.

WEPS+ The World Energy Projection Plus (WEPS+) is a modular system, consisting of separate energy models that are communicate and work together through the overall system model. These models are each developed independently but the modeling system uses a shared historical energy database that allows all the models to communicate with each other when they are run in sequence over a number of iterations. The overall WEPS+ system uses an iterative solution technique that allows for simultaneous convergence of consumption and price to simulate market equilibrium. The WEPS+ model provides forecasts to 2035 of energy consumption and supply for 16 world regions (US, Canada, Mexico and Chile, OECD Europe, Japan, Australia and New Zealand, South Korea, Russia, Other Non-OECD Europe and Eurasia, China, India, Other Non-OECD Asia, Middle East, Africa, Brazil, Other Central and South America). The model is a response surface type of model, used primarily to produce GDP and price responses to a reference case. When a reference case is run, the model calculates the coefficients for the response surface and saves them into a database. When a side case is run, the model uses the previously calculated coefficients to produce a new forecast relative to changes

170

C

IMAGE The Integrated Model to Assess the Global Environment94, 95 (IMAGE) is an ecologicalenvironmental framework that simulates the environmental consequences of human activities worldwide. It represents interactions between society, the biosphere and the climate system to assess sustainability issues like climate change, biodiversity, and human well-being. The objective of IMAGE is to explore the long-term dynamics of global change as the result of interacting demographic, technological, economic, social, cultural, and political factors. IMAGE is a complex modeling framework, consisting of several connected stand-alone software like the TIMER model, the FAIR model, the IMAGE landatmosphere model. Additional to that, IMAGE also uses results from agroeconomic models like LEITAP or IMPACT. Therefore, it is not possible to provide IMAGE as a ‘ready to use package’. The core application of IMAGE is the development and analysis of scenarios of global environmental change. The temporal scale in IMAGE is year based (as for all other model components), although climate parameters (precipitation and temperature) are





generated on a monthly basis through pattern scaling. The number of world regions is 24 (plus Greenland and Antarctica) and historical data for the 1765–2000 period are used to initialize the carbon cycle and climate system, whereas data for 1970–2000 are used to calibrate the energy system (TIMER model) and the agricultural system. IMAGE simulations generally cover the 1970–2050 period and for climate scenarios, often the period 1970–2100. Simulations are made on the basis of scenario assumptions on, for example, demography, food and energy consumption, and technology and trade. The IMAGE model is developed by the IMAGE team under the authority of the Netherlands Environmental Assessment Agency (PBL).

Phoenix (Formerly SGM) Phoenix belongs to the class of dynamic recursive models, and is solved in five-year time steps from 2005 to 2100. Phoenix replaces the Second Generation Model96 (SGM) that was formerly used for general equilibrium analysis at JGCRI. In Phoenix the world is divided into 24 regions: Australia and New Zealand; Brazil; Canada; Central and Other Asia; Central America and Caribbean; China and Taiwan; Eastern Other Europe; EU 15; India; Indonesia; Japan; Korea; Mexico; Middle East; North Africa; Other EU 27; Other Latin America; Rest of World; Russia; South Africa; South Asia; Sub-Saharan Africa; USA; Western Other Europe. Each region includes 26 industrial sectors and two representative agents (the government and a representative consumer). The regional identities include both individual countries (e.g., USA, Brazil, and Canada) and aggregates of countries within a particular geographic region (e.g., Middle East, North Africa, and the EU 15). Each industrial sector produces a single output that is consumed by the representative consumer and the government and used by the production sectors as intermediate inputs. The distribution of goods in each period is based on optimizing behavior by both the producers and consumers; producers minimize costs given a particular nested constant elasticity of substitution (CES) production technology and consumers maximize nested CES utility given a budget constraint. There are four factors of production (natural resources, land, labor, and capital) that are owned by the representative regional household and are supplied to the region’s production sectors. With a given technology, the industrial sectors combine the primary factors of production with intermediate material and energy inputs to produce final consumption


C

goods. When the quantity of goods supplied equals the quantity demanded, the model has achieved market clearance. The flow of goods is accompanied by a corresponding flow of payments. In Phoenix, there are nine electricity-generating technologies (coal, natural gas, oil, biomass, nuclear, hydro, wind, solar, and geothermal) and four additional energy commodities: crude oil, refined oil products, coal, and natural gas. Coal gasification and biomass oil are introduced under the policy scenarios as backstop technologies for the gas and refined oil commodities, respectively. International trade of crude oil, gas, and goods is modeled. A combination of the domestic and imported goods are consumed and used as intermediate inputs as defined by the elasticity of substitution between imported and domestic commodities. The core data are regional social accounting matrices (SAM) generated by the Global Trade Analysis Project (GTAP) version 7 database, which contains detailed economic data for 57 industrial sectors in 112 geographic regions. The GTAP database is supplemented with additional sources, including regional energy data from the International Energy Association (IEA).

MODEL COMPARISON The different models described in the previous section can be used for scenario analysis at the global or local level but each model presents some strengths and weaknesses. The energy analysts, decision makers or in general each user should identify first of all a set of questions that needs an answer from these models. This is really important to identify the most appropriate model case by case in fact each model is different for starting year, time horizon, spatial dimension (number or regions or macroregions) and approach bottom-up (technology rich) or a top-down or hybrid. At the European level, there are only three detailed Member State (MS) energy models: PET, PRIMES, and GEM-E3, whereas all the other models are representing Europe as a single country or sometime split between western and eastern countries. The strengths of the PET model are the number of users, the rich technology database (5 end-use sectors) of Europe and the possibility to easily expand the model to other neighbor regions. The model can be also solved for the 36 ‘European’ regions all together or in a stand-alone version for each single country. Another advantage of this model is the possibility to hard link the TIAM-World model with the


171

Advanced Review


PET European model. This model is useful for energy policy analysis on the medium-long-term horizon of Europe and the linked version TIAM-World-PET can be useful to analyze the security of supply of the European countries. The weaknesses of the model are the missing links between energy and economy. The strength of PRIMES is the modularity of the model with different modules for each demand and supply sector and the formulation based on a microeconomic foundation. The model is demand driven and each sector has a representative decision making agent simulation of markets. Another advantage of the model is the link with the GEM-E3 model and some air quality models for detailed analysis at the local level. The model is used from the European Commission for different studies but at the same time the database is not transparent to other users and not available at all. This is a huge weakness of this model. The GEM-E3 model for Europe is quite different from PET and PRIMES because it is a CGE model of Europe (or the World) that well represents the pricequantity interaction between aggregated sectors. The strength of the model is the possibility to cover the interaction between economy, energy system and environment of the European countries (and/or World Countries). The model can be used to analyze impact on economic growth country by country, projection of employments and unemployment, industrial growth and competition between countries. The weakness of the model is low technological description of some key sectors. The GEMINI-E3, GTEM, MERGE, SGM, and GTAP are all models that can be used to develop scenario analysis similar to GEM-E3. The GEMINI-E3 model presents the same advantages and disadvantages as GEM-E3 but it is focused only on the World level with more regions compared with GEM-E3. The three models GTEM, GTAP, and MERGE can be used to analyze the emission trading mechanisms (clean development mechanism and joint implementation) and the technology market development at the global level. The main weakness of these models regard the low level of description of each aggregated sector in the macroregions. The bottom-up models like TIAM-World (ETSAP-TIAM), MESSAGE, and POLES can be used to analyze the technology development and emission trading mechanism. In this case the sectoral description is more detailed compared with the previous group of models but the interaction between the economic and energy sector is missing. TIAM is an integrated assessment model that can be used to analyze

172

C

the climate-change problem over a time horizon up to 2100. The strength of this model is the very detailed sectoral split and the rich technological database that is necessary to better analyze the climatechange problem but the previously mentioned weakness is still the missing link between economy and energy. In the WITCH model, the hard link between a top-down and bottom-up model is really interesting and useful for some specific analysis. This hybrid model can be used for macroeconomic analysis and technological analysis in the electricity sector. Although compared with other models used for macroeconomic analysis this hybrid approach is quite detailed from the technological point of view, the weakness of this model is still the low technological detail of other sectors. But this is a very good example for future models. Other models to study the climate-change problem are FUND and GCAM. The FUND model is very difficult to use because there is not a user friendly interface. The GCAM model is following a very interesting approach based on the integration of different models but the sectoral split is too simple to well capture the links between energy, environment and economy. The WEPS+ model has a very interesting structure and it is quite detailed but the solution is based on econometric forecast that are very difficult to use on medium and long-term analysis in which the bottomup and top-down approach can generate more interesting results. The simulation approach on which IMAGE is based seems weak compared with the bottom-up and top-down model described before. In addition, the IMAGE model is based on several models that make difficult the use of the entire system. The CIMS model can be used to evaluate the effectiveness and economic impact of public policies to reduce GHG emissions in Canada, USA, and China. The model is used to simulate various policy scenarios and to understand how climate policy affects energy trade. The weakness of the model is due to the multiple assumptions required about energy supply and energy markets to calculate changes in the costs of technologies on the supply side. Another distinction that can be drawn between energy models is the extent of geographic coverage. Local and national models typically are richer in detail than regional or global models but are simplified in terms of international trade. Global models involve significant regional aggregation, the advantages of which relate to:




r

r


the smaller size of the model and hence the lower time required to solve the model, analyze results and build new review/update the model; the lower quantity of information in input and generally less detailed input data;

The disadvantages of regional aggregation however are also important and need to be borne in mind, namely:

r

r

r

the difficulty to analyze behavior at the border between different countries. Generally it is important to analyze and better understand the endogenous trade (import/export) between countries and/or regions (e.g., crossborders electrical interconnections allow the multiregional system to exchange flows between regions in a fully endogenous mode); the regional characterization is less detailed. For example, working with the whole Europe typically involves average data for heating in the residential sector but in reality the energy performance of and energy consumption by buildings in North Europe is quite different from South Europe; the environmental targets that are applied for the whole system and not by country/region.

CONCLUSIONS A number of current bottom-up and top-down energy models that are used at global, regional, or national level were presented and discussed to inform energy analysts in the choice of model to investigate different aspects of the energy system, energy policy, and climate mitigation. These models are used for scenario analysis and various scenarios are modeled to compare the effects of decision making in policy and on business development. This is designed to provide a useful input for decision makers, energy planners, and users to better understand the differences between existing model and approaches. It was ob-

vious that these models even though similar in some regard can be very different in others. Thus the choice of model to study the energy systems is critical and should be fully scoped prior to selecting any existing model. The directions for future research should focus on coupling (through soft or hard linking) between bottom-up and top-down models to better integrate the energy system and the economy, to capture the interactions between energy, the climate, and the economy. In addition, a weakness of the models described in this paper that focus on climate mitigation analysis is the simplified treatment of the agriculture sector. At the local level, for some countries, the contribution of this sector is not small (e.g., agriculture represents 30% of Ireland’s GHG emissions). At the global level this sector is also significant, it is an important contributor to GHG emissions and, when compared with energy, a difficult sector to address in terms of options to reduce emissions. The agriculture sector is deeply connected with the energy system and economic sector and so the future coupling should include energy, economy, climate, and agriculture. The linkages between the energy system, the economy, the climate, agriculture land use, and behavior are all essential elements of robust modeling of future climate scenarios. As this review paper has shown, there are a range of modeling tools and functioning models that focus on energy, the economy and the climate (although with simplified representation of agriculture). The two most underdeveloped areas that require further attention are land use and behavior. Land use provides an interface between agriculture and energy (e.g., the food versus fuel debate). The interface between land use and climate warrants further exploration also. Behavior represents the least understood dimension with macroeconomic models generally assuming rational response to price signals and technoeconomic models assuming agnostic behavioral response to technology change. A key challenge for the future is to develop quantitative methods to incorporate behavioral scientific analysis into energy models.

REFERENCES 1. Bohringer C. The synthesis of bottom-up and top¨ down in energy policy modeling. Energy Econ 1998, 20:233–248. 2. Jebaraj S, Iniyan S. A review of energy models. Renew Sustain Energy Rev 2006, 10:281–311. 3. Greening LA, Bataille C. Bottom-up models of energy: across the spectrum. In: Evans J, Hunt LC, eds. In-


C

ternational Handbook on the Economics of Energy (chapter11). Cheltenham, UK and Northampton, MA: Edward Elgar; 2009. ´ Gallachoir 4. Deane JP, Chiodi A, Gargiulo M, O ´ B. Soft-linking of a power systems model to an energy systems model. Energy Policy 2012, 42:303– 312.


173

Advanced Review


5. Frei C, Haldi P, Sarlos G. Dynamic formulation of a top-down and bottom-up merging energy policy model. Energy Policy 2003, 31:1017–1031. 6. Horne M, Jaccard M, Tiedemann K. Improving behavioral realism in hybrid energy-economy models using discrete choice studies of personal transportation decisions. Energy Econ 2005, 27:59–77. 7. Bauer N, Edenhofer O, Kypreos S. Linking energy system and macroeconomic growth models. Comput Manag Sci 2008, 5:95–117. 8. Nakichenovich N. et al. Climate change 1995, Impacts, Adaptation and Mitigation of Climate Change: Scientific–Technical Analyses, Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel for Climate Change, WMO– UNEP. Cambridge, UK: Cambridge University Press; 1996, 75–78. 9. Weyant JP. Policy modeling: an overview. Energy 1990, 15:203–206. 10. Capros P, Karadeloglou P, Mentzas G, Samouilidis JE. Short and medium-term modelling and problems of models linkage. Energy 1990, 15:301–324. 11. Lehtila A, Pirila P. Reducing energy related emissions: using an energy systems optimization model to support policy planning in Finland. Energy Policy 1996, 24:805–819. 12. Messner S, Schrattenholzer L. MESSAGE-MACRO: linking an energy supply model with a macroeconomic module and solving it iteratively. Energy 2000, 25:267–282. 13. Yamamoto H, Yamaji K, Fujino J. Scenario analysis of bioenergy resources and CO2 emissions with a global land use and energy model. Appl Energy 2000, 66:325–337. 14. Argonne National Laboratory. Greenhouse Gas Mitigation Analysis Using ENPEP: A Modeling Guide. Vienna, Austria: International Atomic Energy Agency; 2001. 15. Babiker M, Reilly J, Mayer M, Eckaus R, Wing I, Hyman R. The MIT Emissions Prediction and Policy Analysis Model: Revisions, Sensitivities, and Comparisons of Results. Cambridge, MA: MIT Joint. Program on the Science and Policy of Global Change, Massachusetts Institute of Technology; 2001. 16. Barreto L, Kypreos S. Multi-regional technological learning in the energy-systems MARKAL model. Global Energy Issues 2002, 17:189–213. 17. Labriet M, Loulou R, Kanudia A. Global energy and CO2 emission scenarios—analysis with a 15 region world MARKAL model. In: Haurie A, Viguier L, eds. The Coupling of Climate and Economic Dynamics. Boston/Dordrecht/London: Kluwer Academic; 2004, 205–235. 18. Bataille C, Jaccard M, Nyboer J, Rivers N. Towards general equilibrium in a technology-rich model with empirically estimated behavioral parameters. Energy

174

C

J. 2006, 27 (Special Issue on Hybrid Modeling of Energy–Environment Policies: Reconciling Bottom-up and Top-down):93–112. 19. Baumann M. EFDA-TIMES Model: management, coordination and documentation of the Model Master Version. Final Report. SubTask TW6-TREETMMAN. SERF7. April, 2008. 20. Howells M, Rogner H, Strachan N, Heaps C, Huntington H, Kypreos S, Hughes A, Silveira S, DeCarolis J, Bazillian M, Roehrl A. OSeMOSYS: The Open Source Energy Modeling System. An introduction to its ethos, structure and development. Energy Policy 2011, 39:5850–5870. ˚ Carson M, Pachauri 21. Nilsson M, Heaps C, Persson A, S, Kok M, Olsson M, Rehman I, Schaeffer R, Wood D, Van Vuuren D, Riahi K, Americano B, Mulugetta Y. Energy for a Shared Development Agenda: Global Scenarios and Governance Implications. SEI Report; 2012. 22. Uri ND. Forecasting peak system load using a combined time series and econometric model. Appl Energy 1978, 4:219–227. 23. Uri ND, Flanagan SP. Short-term forecasting of crude petroleum and natural gas production. Appl Energy 1979, 5:297–310. 24. Jeng-Wen L, Chia-Yon C. A cost minimization model for coal import strategy. Energy Policy 1996, 24:1111– 1117. 25. Xie Z, Kuby M. Supply-side-demand side optimization and cost-environment trade-offs for China’s coal and electricity system. Energy Policy 1997, 25:313– 326. 26. Henning D. MODEST—an energy system optimization model applicable to local utilities and countries. Energy 1997, 22:1135–1150. 27. Steckley SG, Meade DS, Lenox CS, Hoffman KC, Reid DH, Schoener BCH. Energy demand analytics using coupled technological and economic models. Energy J 2011, 32 (Special Issue I, Strategies for mitigating climate change through energy efficiency: a multi-model perspective):173–192. 28. Energy Modeling Forum (EMF). EMF 25 Scenario Design. Stanford, CA: Stanford University; 2010. Available at: http://emf.stanford.edu/files/res/ 2372/25design2.pdf. (Accessed November 2012). 29. Energy Modeling Forum (EMF). Available at: http://emf.stanford.edu/docs/about_emf/. (Accessed November 2012). 30. Clarke L, Edmonds J, Krey V, Richels R, Rose S, Tavoni M. International climate policy architectures: overview of the EMF 22 International Scenarios. Energy Econ 2009, 31:S64–S81. 31. IIASA-ECS models description. Available at: http:// www.iiasa.ac.at/Research/ECS/docs/models.html# MESSAGE. (Accessed November 2012).





32. Messner S, Strubegger M. User’s Guide for MESSAGE III, WP-95–69. International Institute for Applied Systems Analysis 1995, Laxenburg, Austria. 33. Riahi K, Roehrl RA. Energy technology strategies for carbon dioxide mitigation and sustainable development. Environ Econ Policy Stud 2000, 3:89–124. 34. Klaassen G, Riahi K. Internalizing externalities of electricity generation: an analysis with MESSAGE– MACRO. Energy Policy 2007, 35:815–827. 35. MESSAGE Regions Definition. Available at: http:// www.iiasa.ac.at/Research/ECS/docs/11worldregions. html. (Accessed November 2012). 36. Pan European TIMES Model. Available at: http:// www.kanors.com / DCM / PET36 / Docs / Index.aspx. (Accessed November 2012). 37. Loulou R, Remme U, Kanudia A, Lehtila¨ A, Goldstein G. Documentation for the TIMES Model. Energy Technology Systems Ananlysis Programme (ETSAP), April 2005. 38. IEA-ETSAP. Available at: http://www.iea-etsap.org/. (Accessed November 2012). 39. New Energy Externalities Development for Sustainability—FP6 Project. Available at: http://www. needs-project.org/. (Accessed November 2012). 40. Monitoring and Evaluation of the RES Directives Implementation in EU27 and Policy Recommendations for 2020–FP7 Project. Available at: http:// www.cres.gr/res2020. (Accessed November 2012). 41. Risk of Energy Availability: Common Corridors for Europe Supply Security–FP7 Project. Available at: http://reaccess.epu.ntua.gr. (Accessed November 2012). 42. Research, methodologies and technologies for the effective development of Pan-European key grid infrastructures to support achievement of a reliable, competitive and sustainable electricity supply–FP7 project. Available at: http://realisegrid.rse-web.it/. (Accessed November 2012). 43. Irish-TIMES Model at University College Cork. Available at: http://www.ucc.ie/en/serg/energypolicy/ projects/irishtimes/. (Accessed November 2012). 44. Capros P, Mantzos L, Papandreou V, Tasios N. European Energy and Transport. Trends to 2030 (Update 2007). European Commission; 2008. 45. Capros P. The PRIMES Energy System Model: Summary Description. National Technical University of Athens–European Commission Joule-III Programme. 46. Capros P. The E3MLab Models. Presented at the PRIMES Model Seminar; Prague, Czech; January 17, 2011. 47. Tubbs WJ. A Simulation Model of Energy Supply and Demand for Climate Policy Analysis. Canada: Master of Resource Management Simon Fraser University Burnaby, BC; 2008. 48. Xianqiang M, Youkai X, Shengqiang L, Qin L. The application of CIMS model in China. Asia Energy


C

49.

50.

51.

52.

53.

54.

55.

56. 57.

58.

59. 60.

61.

62.

63.

64.

Environment Modeling Forum (AEEMF); 2008. Available at: http://home.hiroshima-u.ac.jp/aeemf/ 4th_meeting / 4th_materials / 3rdday / S11_32_Liu.pdf. (Accessed November 2012). Bosetti V, Massetti E, Tavoni M. The WITCH model: structure, baseline, solutions. FEEM Working paper N.10 2007. Bosetti V, Carraro C, Galeotti M, Massetti E, Tavoni M. WITCH: a world induced technical change hybrid model. Energy J, Special Issue “Hybrid Modelling of Energy Environment Policies: Reconciling Bottom-up and Top-down”; 2006, 13–38. Bosetti V, Carraro C, Tavoni M. Timing of mitigation and technology availability in achieving a lowcarbon world. Environ Resour Econ 2012, 51:353– 369. Loulou R, Labriet M. ETSAP-TIAM: the TIMES integrated assessment model Part I: model structure. Computational management science, special issue. Manag Energy Environ 2008, 5:7–40. Loulou R. ETSAP-TIAM: the TIMES integrated assessment model Part II: mathematical formulation. Computational management science, special issue. Manag Energy Environ 2008, 5:41–66. Loulou R. Regional economic and energy implications of reaching global climate targets—A policy scenario analysis. Deliverable n.12 PLANTES project. 2009. Babonneau F, Haurie A, Loulou R, Vielle M. Combining stochastic optimization and Monte Carlo simulation to deal with uncertainties in climate policy assessment. Environ Model Assess 2012, 17:51–76. POLES model description—http://www.enerdata.net/ enerdatauk/solutions/energy-models/poles-model.php. World energy, technology and climate policy outlook 2030, WETO. Luxembourg: Office for Official Publications of the European Communities; 2003. World energy technology outlook 2050, WETO-H2. Luxembourg: Office for Official Publications of the European Communities; 2006. Deciding the future: energy policy scenarios to 2050. London, UK: World Energy Council; 2007. GTAP Model Documentations. Available at: https:// www.gtap.agecon.purdue.edu/models/current.asp. (Accessed November 2012). Burniaux JM, Troung P. GTAP-E: an energyenvironmental version of the GTAP model. GTAP Technical Paper n.16; 2002. Bernard A, Drouet L, Vielle M. GEMINI-E3, A General Equilibrium Model of International-National Interactions between Economy, Energy and the Environment, V5.2. D3.3 TOCSIN Project no. 044287; 2008. Bernard A, Vielle M. Assessment of European Union transition scenarios with a special focus on the issue of carbon leakage. Energy Econ 2009, 31:S274–S284. Bernard A, Vielle M. Report to French Inter Ministerial Task Forces on Climate Change: Evaluations of the


175

Advanced Review

65.

66.

67.

68.

69.

70.

71.

72. 73.

74.

75.

76.

77.

176


Kyoto Protocol with GEMINI-E3. Technical Report; 1999. Capros P. et al. GEM-E3: Computable General Equilibrium Model for Studying Economy–Energy– Environmental Interactions. Brussels: European Commission Report EUR 16714 EN; 1995. Capros P. et al. The GEM-E3 Model: Reference Manual. Athens: National Technical University of Athens Report; 1997. Capros P, Georgakopoulos T, Van Regemorter D, Schmidt T, Conrad K. The GEM-E3 General Equilibrium Model for the European Union. International Symposium in Economic Modeling, Vienna, 1998. J Econ Finan Model 1998, 51–160. Maisonnave H, Pycroft J, Saveyn B, Ciscar MJ. Does climate policy make the EU economy more resilient to oil price rises. A CGE analysis. Energy Policy 2012, 47:172–179. Saveyn B, Paroussos L, Ciscar JC. Economic analysis of a low carbon path to 2050: A case for China, India and Japan. Energy Econ 2012, 34:S451– S458. GEM-E3 Documentation. Available at: http://ipts.jrc. ec.europa.eu/activities/energy-and-transport/gem-e3. (Accessed November 2012). GTEM Model Description. Available at: http://www. daff.gov.au/abares/models. (Accessed November 2012). ABARE. The MEGABARE model: interim documentation. Canberra: ABARE; 1996. Ahammad H, Matysek A, Fisher BS, Curtotti R, Gurney A, Jakeman G, Heyhoe E, Gunasekera D. Economic impact of climate change policy: the role of technology and economic instruments. ABARE Research Report 06.7. Canberra: ABARE; 2006. Ahammad H, Mi R. Land Use Change Modeling in GTEM: Accounting for forest sinks. EMF22: Climate Change Control Scenarios, Tresider Memorial Union. California: Stanford University; 2005. Global Change Assessment Model Description. Available at: http://www.globalchange.umd.edu/ models/gcam/. (Accessed November 2012). Edmonds J, Wise M, Pitcher H, Richels R, Wigley T, MacCracken C. An Integrated Assessment of Climate Change and the Accelerated Introduction of Advanced Energy Technologies: An Application of MiniCAM 1.0. Mitigat Adapt Strat Global Change 1996, 1:311– 339. Edmonds J, Wise M, Sands R, Brown R, Kheshgi H. Agriculture, Land-Use, and Commercial Biomass Energy: A Preliminary Integrated Analysis of the Potential Role of Biomass Energy for Reducing Future Greenhouse Related Emissions. Washington, DC: PNNL-11155, Pacific Northwest National Laboratories; 1996.

C

78. Wigley TML, Raper SCB Future changes in globalmean temperatue and sea level. In: Climate and Sea Level Change: Observations, Projections and Implications. Cambridge, UK: Cambridge University Press; 1993, 111–133. 79. Hulme M, Jiang T, Wigley T. SCENGEN: A Climate Change SCENario GENerator: Software User Manual, Version 1.0. Climate Change Research Unit, School of Environmental Sciences. Norwich, UK: University of East Anglia; 1995, 38. 80. Kim SH, Edmonds JA, Lurz J, Smith SJ, Wise M. The ObjECTS framework for integrated assessment: hybrid modeling of transportation. Energy J 2006, 27:63–91. 81. Kyle GP, Clarke LE, Rong F, Smith SJ. Climate policy and the long-term evolution of the U.S. buildings sector. Energy J 2010, 31:145–172. 82. FUND Model. Available at: http://www.fundmodel.org/. (Accessed November 2012). 83. Link PM, Tol RSJ. The economic impact of a shutdown of the thermohaline circulation: an application of FUND. Climatic Change 2011, 104:287–304. 84. Tol RSJ. On the optimal control of carbon dioxide emissions: an application of FUND. Environ Model Assess 1997, 2:151–163. 85. Anthoff D, Tol RSJ. The Uncertainty about the Social Cost of Carbon: A Decomposition Analysis Using FUND. ESRI Working Paper N.404; 2011. 86. Bohringer C, Rutherford TF, Tol RSJ. The EU ¨ 20/20/2020 targets: an overview of the EMF22 assessment. Energy Econ 2009, 31:S268–S273. 87. MERGE Model Description. Available at: http://www.stanford.edu/group/MERGE/. (Accessed November 2012). 88. Manne A, Mendelsohn R, Richels R. MERGE: a model for evaluating regional and global effects of GHG reduction policies. Energy Policy 1995, 23:17–34. 89. LEAP Documentation Page. Available at: http://www. energycommunity.org/default.asp?action=47. (Accessed November 2012). 90. LEAP Applications. Available at: http://www. energycommunity.org/default.asp?action=45. (Accessed November 2012). 91. Heaps C, Erickson P, Kartha S, Kemp-Benedict E. Europe’s Share of the Climate Challenge Report published by SEI in partnership with Friends of the Earth Europe; 2009. 92. World Energy Projection System Plus Model Documentation 2010. World Industrial Model. U.S. Energy Information Administration. Washington, DC: U.S. Department of Energy-M079; 2010. 93. The World Energy Projection System Plus (WEPS+): An Overview 2011. Energy Information Administration. Washington, DC: U.S. Department of EnergyM074; 2011.





94. Integrated Model to Assess the Global Environment Description. Available at: http://themasites.pbl.nl/ en/themasites/image/index.html. (Accessed November 2012). 95. Tomassini L, Knutti R, Plattner G, Van Vuuren DP, Stocker TF, Howarth RB, Borsuk ME. Uncertainty


C

and risk in climate projections for the 21st century. Climatic Change 2011, 103:399–422. 96. Pitcher HM, Sands RD, Fawcett AA. The Second generation model: future directions for model development. PNNL-15437. College Park, MD: Joint Global Research Institute.


177