Deliverable D - UoA...PROMITHEAS-4: “Knowledge transfer and research needs for preparing...

PROMITHEAS-4: “Knowledge transfer and research needs

for preparing mitigation/adaptation policy portfolios”

Deliverable D.3

“Choice and implementation of Models for

Mitigation / Adaptation policy portfolios”

Lead Beneficiary

Included Overviews

Overview of Models in use for Mitigation/Adaptation policy

Selection of Models for Mitigation / Adaptation policy



PPRROOMMIITTHHEEAASS –– 44

Task Leader: Prof. Bernhard Felderer,

Institute of Advanced Studies (IHS),

Vienna, August 2011

““OOvveerrvviieeww ooff MMooddeellss iinn UUssee ffoorr

MMiittiiggaattiioonn//AAddaappttaattiioonn PPoolliiccyy””



This report has been read, commented and approved by all members of the PROMITHEAS-4 Scientific Committee. It was also disseminated for comments, through BSEC – PERMIS and BSEC – BC, to all relevant governmental and business authorities and partners before its finalization. Partners from the beneficiary countries* of the consortium were encouraged to contact direct national authorities, agencies, institutions and market stakeholder for comments before the finalization of this report (Annex 1). Members of the PROMITHEAS – 4 Scientific Committee:

1. Prof. Dimitrios MAVRAKIS, NKUA – KEPA (GREECE) -Editor 2. Dr. Popi KONIDARI, NKUA – KEPA (GREECE) – Assistant to the editor 3. Dr. Harry KAMBEZIDIS, NOA (GREECE) 4. Prof. Bernhard FELDERER, IHS (AUSTRIA) 5. Prof. Bilgin HILMIOGLU, TUBITAK – MAM (TURKEY) 6. Prof. Vahan SARGSYAN, SRIE – ESC (ARMENIA) 7. Prof. Dejan IVEZIC, UB – FMG (SERBIA) 8. Prof. Mihail CHIORSAK, IPE ASM (MOLDOVA) 9. Prof. Agis PAPADOPOULOS, AUT – LHTEE (GREECE) 10. Prof. Alexander ILYINSKY, FA (RUSSIA) 11. Prof. Anca POPESCU, ISPE (ROMANIA) 12. Prof. Andonaq LAMANI, PUT (ALBANIA) 13. Prof. Elmira RAMAZANOVA, GPOGC (AZERBAIJAN) 14. Dr. Lulin RADULOV, BSREC (BULGARIA) 15. Prof. Arthur PRAKHOVNIK, ESEMI (UKRAINE) 16. Prof. Sergey INYUTIN, SRC KAZHIMINVEST (KAZAKHSTAN) 17. Prof. Alvina REIHAN, TUT (ESTONIA)

*Turkey, Armenia, Serbia, Moldova, Russia, Romania, Albania, Azerbaijan, Bulgaria, Ukraine,

Kazakhstan, Estonia.

The EU, the Consortium of PROMITHEAS – 4 and the members of the Scientific Committee do not undertake any

responsibility for copyrights of any kind of material used by the Task Leaders in their report. The responsibility is

fully and exclusively of the Task Leader and the his/her Institution.

Acknowledgments: The Task Leader of this report acknowledges the contribution of Mr. Michael-Gregor Miess and Mr. Stefan Schmelzer for the development of this overview.

PROMITHEAS-4: “Knowledge transfer and research needs for

preparing mitigation/adaptation policy portfolios”

2

Table of Contents Table of Abbreviations 3

Introduction 4

Integrated Scenarios 6

Integrated Assessment Models for Adaptation/Mitigation 8

MARKAL/TIMES 12

Specific Characteristics of MARKAL 12

Specific Characteristics of TIMES 14

Evaluation of MARKAL/TIMES 14

ENPEP-BALANCE 16

Specific Characteristics of ENPEP-BALANCE 16

Evaluation of ENPEP-BALANCE 17

MESSAGE 17

Specific Characteristics of MESSAGE 18

Evaluation of MESSAGE 19

LEAP 21

Specific Characteristics of LEAP 21

Evaluation of LEAP 22

IMAGE 24

Specific Characteristics of IMAGE 24

Evaluation of IMAGE 25

MERCI 25

Specific Characteristics of MERCI 26

Evaluation of MERCI 27

Conclusion 27

References 29



Table of Abbreviations

ADAM Title of the “Adaptation and Mitigation Strategies: Supporting European Climate Policy” project BAU Business As Usual CCS Carbon Capture and Storage CEEESA Centre for Energy, Environmental and Economic Systems Analysis CES Constant Elasticity of Substitution EARDF European Agricultural Rural Development Fund EC European Commission EFOM Energy Flow Optimization Model ENPEP Energy and Power Evaluation program ERDF European Regional Development Fund ESF European Social Fund ETSAP Energy Technology and Systems Analysis Program FAIR Framework to Assess International Regimes for differentiation of commitments GAMS General Algebraic Modeling System GDP Gross Domestic Product GHG GreenHouse Gas IAEA International Atomic Energy Agency IAM Integrated Assessment Model IHS Institut für Höhere Studien (Institute for Advanced Studies) IIASA International Institute for Applied Systems Analysis IEA International Energy Agency IMAGE Integrated Model to Assess the Global Environment IPCC Intergovernmental Panel on Climate Change LEAP Long-range Energy Alternatives Planning M/A Mitigation/Adaptation MARKAL Market Allocation (Model) MERCI Model for Evaluating Regional Climate change Impacts MESSAGE Model for Energy Supply Strategy Alternatives and their General Environmental (Impact) MNP Milieu en Natuur Planbureau (Netherlands Environmental Assessment Agency) NAPA National Adaptation Programs for Action OECD Organisation for Economic Co-operation and Development R&D Research and Development RES Reference Energy Scenario RES Renewable Energy Sources RIVM Rijksinstituut voor Volksgezondheid en Milieu (National Institute for Health and Environment) SEI Stockholm Environment Institute TED Technology and Environmental Database TIMES The Integrated MARKAL-EFOM System UNFCCC United Nations Framework Convention on Climate Change WEM World Energy Model



4

Introduction

This report provides an overview of models that should be considered to be used for developing and quantifying adaptation and mitigation scenarios for the countries of the Black Sea Region. These models are evaluated according to the specific aspects to be regarded for the emerging economies of the countries in question, according to the advantages or disadvantages when using them for scientific and public purposes, as well as according to their terms and costs of use. The challenges of developing a stable and sustainable energy system are manifold. Research in the industrial countries was conducted on a very broad range, however, it still has to be clarified if and how these research results can be applied to the situation in emerging economies. This is the focus of the PROMITHEAS-4 project: to develop and evaluate mitigation and adaptation policy portfolios, together with a characterization of research needs and gaps in this area. Countries of the Black Sea Region are the predominantly targeted emerging economies of this project: Albania, Armenia, Azerbaijan, Bulgaria, Moldova, Romania, Russian Federation, Serbia, Turkey and Ukraine. Estonia and Kazakhstan are also included in the beneficiary countries of this project. Their economic characteristics, however, are comparable to the ones of the Black Sea Region. The assessment of the main characteristics of emerging economies in this region will be crucial for the suitability of energy models for the use in generating M/A policy portfolios in these countries. As Urban F. et al (2007) have elaborated, there is reason to question the use of energy models developed in/for industrialised countries in developing countries. They find that the characteristics of the energy systems and economies of developing countries differ from those in industrialised countries in the following aspects, which apply to developing countries’ energy systems:

• The electricity supply of the economy is not functioning optimally;

• The electrification rates are much lower relative to industrialized countries;

• Predominant use of traditional bio-fuels;

• The tariffs are often below long-term marginal cost of production and many bills will

never be paid;

• A widespread informal economy;

• Developing countries may not follow the same trajectory towards industrialisation as

today’s industrialized countries did;

• The urban-rural divide causing high distribution differences within countries and regions;

• Abuse or inadequate use of subsidies;

(Urban F. et al. 2007, p. 3474 ff). Data from “Procedures, sources, and data for Mitigation / Adaptation for policy portfolios” report could be used to clarify the following question: What exactly are the specific requirements for emerging economies? There is little literature dealing with this specific issue. The situation in the Black Sea Region is of course a different one to that of developing countries. However, it is important to bear in mind that energy models in use for industrialised countries may yield insufficient or misleading outcomes for emerging economies. Therefore, a short description of the economic situation in the Black Sea Region is important for the comparison of different energy modelling tools. After defining the characteristics the models should incorporate, we will analyse them from this perspective.



5

A concise description of the current economic situation in the Black Sea Region can be found in a policy report prepared by the Commission on the Black Sea (Gavras, 2010). After 1999 the region experienced a period of high growth rates, growing government credibility, improvement of the legal system, low government deficits, and rapid movement towards a well functioning market-oriented economic system. However, the financial crisis of 2008 brought this upward trend to a halt. Especially the large decrease in investment inflows into the region caused serious problems, although the situation differs among the countries. Obtaining funds became difficult, since investment into the region’s countries is perceived to carry higher risks and risk aversion is reaching high levels during the current financial crisis. Through lower credit ratings (there are other factors determining a country’s risk as well, but this is the most prominent one) the cost of financing budget deficits and undertaking investments rises. Furthermore, the economies have little possibilities to conduct effective stimulus programmes (Gavras p. 7ff, p. 13ff). After having provided a very brief sketch of the economic framework within which any model employed by the PROMITHEAS-4 project will operate, we have to look at what kinds of models are eligible for use. The final set of models included in this study consist of the following models: MARKAL/TIMES, ENPEP-BALANCE, MESSAGE, LEAP, IMAGE and MERCI. A wide spectrum of models has been developed over the past thirty (30) years, which by far cannot be covered within this project. We therefore selected the most renowned models that, in advance, were considered to be best suited for the PROMITHEAS-4 project. Moreover, the set of models is such that at least one simulation, economic equilibrium or optimisation1 model is included. PRIMES, for example, which is also a renowned simulation model, is not included because it is similar to ENPEP-BALANCE. Similarly, EnergyPLAN is also a simulation model, which was not included, since otherwise the simulation models would be overrepresented. Important for the inclusion of a model within our study was its suitability for emerging economies, such as evolved by Bhattacharyya and Timilsina (2010), as well as by Urban et al. (2007). E.g. PRIMES and EnergyPLAN are not mentioned within these studies. Another considered criterion was the availability of model applications on a national level and moreover, in countries of the Black Sea region (e.g. PRIMES is predominantly used at European level and databases are not available for many countries; the World Energy Model (WEM), although mentioned by Urban et al. (2007), is mainly employed at a global scale and therefore not included herein). Summarising, there is a vast number of models that could be included within this overview. However, we decided to choose the most representative ones for each modelling category (simulation, optimisation and economic equilibrium), as well as those with the highest probability to be suited for emerging economies within the Black Sea region. As the selected model will be used to design integrated M/A scenario portfolios, comprising environmental, technological, economic and policy problems, we have to focus on integrated assessment models that are able to deliver this purpose. Thus, before we can begin with a modelling overview, we have to look at the requirements these models have to fulfil, i.e. what we understand as integrated scenarios.

1 These categories will be explained in more detail in the next two sections.



6

Integrated Scenarios

As it is elaborated on in the final report of the ADAM (Adaptation and Mitigation Strategies:

Supporting European Climate Policy) project, adaptation and mitigation, even though some consider them as alternative strategies to deal with climate change, should not be regarded as mutually exclusive (Hulme et al., p. 8). Thus, any policy portfolio designed for the PROMITHEAS-4 project should involve mitigation as well as adaptation strategies, not only trying to limit the global temperature increase to 2°C (which would involve already a high level of mitigation activities, see Hulme et al., p. 8), but also include measures to deal with the effects of this temperature increase or a larger one of up to 4°C. Generally, a scenario should be seen as time paths of key variables, which are exogenously specified and then used as inputs driving other parts of an assessment (Parson and Fisher-Vanden, 1997, p. 595). Viewed in the context of energy-economy-environmental modelling, then, scenarios are story-lines about how an energy system might evolve over time2. However, an M/A scenario will not only have to consider the development of the energy system and the emission patterns it induces (mitigation), but also take into account the regional aspects of adaptive measures. Therefore, while mitigation, seen as identification of the appropriate set of energy efficient and renewable energy technologies, has been intensively studied and has a common knowledge base, e.g. a unified technological database, defined expectations for technological progress, etc., adaptation includes responses to the predicted impact of climate change and is highly dependent on local conditions and on national priorities. Thus, any modelling effort has to be supported by a model that is flexible enough to consider regional, local, and national aspects of adapting to climate change. As has emerged from the recent discussion on climate change policies (both at European and global level), adaptation represents a new priority for policy makers, which also has to be taken into account when constructing policy portfolios. This requires not only a change in attitude, but also a change in the scale of priorities for both policy makers and scientists. In a complex financial framework such as that present for the European Union, a change in the scale of priorities, to be effective, must correspond to a shift of budgetary allocations. This step has high-level political implications and therefore needs to be dealt with in a wider framework which critically evaluates all current EU policies and priorities (e.g. cohesion, competitiveness, growth, infrastructure, etc.) (Lavalle, 2009, p. 4). In order to take on this matter, the European Commission has issued a White Paper, “Adapting to climate change: Towards a European framework for action” (EC White Paper in the following) in 2009 as its latest formal proposal. This White Paper is based on a phased approach, the first of which will last for the period 2009-2012, preparing a comprehensive EU adaptation strategy, which is then to be implemented in phase 2 starting from 2013 (see EC White Paper, p. 7). Phase 1 is based on four (4) main pillars (EC White Paper, p. 7):

1) building a solid knowledge base on the impact and consequences of climate change for the

EU;

2) integrating adaptation into EU key policy areas;

3) employing a combination of policy instruments to ensure effective delivery of adaptation

and;

2 UNFCCC “Module 5.1 – Mitigation Methods and Tools in the Energy Sector” 2006:55



7

4) stepping up international cooperation on adaptation.

In order for this phase 1 to work, national, regional and local authorities have to cooperate closely. The EU provides funding for financing adaptation measures, amongst other through the following sources:

• The Common Agricultural Policy and Rural Development (EARDF);

• The Structural Fund (ERDF);

• The European Solidarity Fund (ESF);

• Civil Protection Mechanism.

To encourage developing countries and emerging economies to prepare for the expected impacts of climate change, the United Nations Framework Convention on Climate Change (UNFCCC) proposes to develop National Adaptation Programs for Action (NAPA)3, including information, among other things, on climate change induced natural hazards like floods, droughts, heat waves, heavy rainfall, hurricanes, and tornadoes. Even though the NAPAs are rather designed for less developed countries, which stand in contrast to the emerging economies of the Black Sea region, it seems clear that any adaptation strategy is subject to a larger framework at the EU and global level, which is being developed at this very moment. All of these facts point to the conclusion that only a selected number of Integrated Assessment Models for Adaptation/Mitigation (IAMs) are eligible for use by the PROMITHEAS-4 project. However, one has to say that most models primarily focus on mitigation issues, and that adaptation, if present in the models, can only be implicitly depicted in the majority of cases. Therefore, adaptation issues probably will have to be dealt with also outside of a formal modelling environment, defining the adaptation part of an M/A policy portfolio predominantly based on regional economic and environmental specifics. After an introduction into the class of IAMs, a selection of those that come into question for the PROMITHEAS-4 project will be evaluated.

3 http://unfccc.int/national_reports/napa/items/2719.php



8

Integrated Assessment Models for Adaptation/Mitigation

Before one can talk of an assessment model, it is crucial to define this term. If one follows the literature, assessment is described as

“social processes that bridge the domains of knowledge and decision-making, assembling and synthesizinig expert scientific or technical knowledge to advise policy or decision-making”.

(Parson, Fisher-Vanden, 1997, p. 590, see also Parson, 1995, and Weyant et al., 1996)

One way to achieve such an assessment is by employing a formal modelling environment that represents the complex relationships underlying the to-be-assessed problem field, as opposed to e.g. deliberation by interdisciplinary expert panels (see Parson, Fisher-Vanden, 1997, p. 591). The Promitheas-4 project has chosen to rely mainly on a formal modelling environment to construct M/A policy portfolios for the beneficiary countries.

Integrated assessment models, now, can deliver this purpose by combining socio-economic dimensions of climate change with systemic aspects of technological alternatives in order to address policy options and environmental impacts of climate change. In general, IAMs attempt to employ one or several of three methods associated with each other, in a combined or stand-alone form, to project emissions and with them climate change (Parson, Fisher-Vanden, 1997, p. 595): emission scenarios (externally specified), an accurate bottom-up

representation of technologies for the production of energy and other goods, and economic

modelling in an aggregate form (e.g. taking account of economic equilibrium conditions). Common to all these approaches are predictions of the future in a speculative form, who only “differ in the detail and explicitness of different components of the projections” (Parson, Fisher-Vanden, 1997, p. 595). Given the speculative nature of these models, they serve as a means to estimate costs and benefits of policy options, always related to a possible future development of the social, economical and environmental system, all of them being dependent on each other to a certain extent. Hence, they can be used to (Dickinson, 2007, p.7):

• Assess climate change control policies (Weyant et al., 1996);

• Create interdisciplinary frameworks;

• Address climate change problems including determining influential forces that make

sectors sensitive to climate change;

• Quantify environmental and non-environmental problems resulting from climate

change by ranking climate change control benefits and detriments in developed and

emerging economies, as well as developing countries (IPCC, 2001).

IAMs can be divided into three (3) types of models: simulation, economic equilibrium and optimisation models (see e.g. Urban F. et al., p.3479). However, Bhattacharyya and Timilsina (2010) use a different type of categorisation, especially designed for energy system models (see Bhattacharyya, p. 501):

• bottom-up, optimisation-based models (such as e.g. MARKAL);

• bottom-up, accounting models (such as e.g. LEAP);



9

• top-down, econometric models;

• hybrid models (such as e.g. MERCI);

• electricity system models.

Which categorisation is used will depend on the characteristics one wants to include in the differentiation of the models. For the purposes of this report, the categorisation from Urban F. et al. (2007), which essentially is an adaptation of a model classification system proposed by Van Beeck (1999,2003) and not much different to the one of Bhattacharyya and Timilsina (2010), has been chosen as the most suitable. The different aspects to and specifics of the different model classes are described in brief in Table 1 below.

Table 1: Integrated Assessment Models Comparison Chart

MODEL Simulation IAMs Economic Equilibrium

IAMs Optimisation IAMs

Use

To simulate and approximate the

environmental results of a selected policy

option

To assess overall economic development and ecological impacts

simultaneously

To identify optimal policies e.g. of climate change

control options

Scale Individual portfolio of policy options or

chosen scenarios Global or national level Usually global

Description

Through the use of scenarios based on

user-defined assumptions, a

portfolio of policy options is produced

Finds a new economic equilibrium based on exogenously specified

scenarios, endogenously finding optimal control

variables

Determines the policy path that maximizes utility, or minimizes costs, while imitating the effects of

mitigation on the global/local economy

Examples LEAP, IMAGE,

ENPEP MERCI

MARKAL/TIMES, MESSAGE

Source: Dickinson, 2007, p.8, Urban F. et al, p. 3479, Authors. Integrated Assessment models were first designed beginning from the 1990’s, when the focus of policy makers and the scientific community shifted towards energy-environment interactions and climate change related issues (Bhattacharyya and Timilsina, 2010, p. 498). To depict these environmental issues, some extensions to energy system models were close at hand (Bhattacharyya and Timilsina, 2019, p. 498):

• accounting models, i.e. models based on energy balances (see Bhattacharyy and Timilsina,

2010, p. 496), were able to incorporate environmental effects in relation to energy

production, conversion and use by including an appropriate set of environmental

coefficients;

• network-based models, i.e. models extending the energy balance framework to a network

description of the energy system capturing all activities involved in the entire supply chain

(see Bhattacharyy and Timilsina 2010, p. 496), could similarly estimate environmental

burdens employing environmental pollution coefficients and evaluating the economic

impacts by considering costs of mitigation;

• energy models with macro linkage could analyse the allocation issues taking account of the

overall economic implications.



10

Bhattacharyya and Timilsina (2010) come to the conclusion that what they call models of bottom-up accounting type (or simulation IAMs, in the classification of this report) are best suited for the representation of energy systems for developing countries, mostly because of their flexibility, limited skill requirements (Bhattacharyya and Timilsina, 2010, p. 501), and because they can account for several specifics of developing countries. Under the conditions delineated above, potential integrated scenario-based assessment models that include mitigation and adaptation were considered as a result of an extensive literature search, parts of which are presented in the next sessions and after communications with key Partners during the kick-off meeting, 3rd-4th March 2011, Athens.

Figure 1: Criteria for Models to be Included in this Overview

Source: Authors.

The models subsequently described in this overview are presented in Table 2 below. The following overview is based on (a) model documentations, (b) existing research reports, (c) articles in scientific journals and (d) model websites.

PROMITHEAS-4: “Knowledge transfer and research needs for preparing mitigation/adaptation policy

portfolios”

Table 2: List of Models to be surveyed

MODEL NAME Organisation / Author FURTHER INFORMATION Availability

MARKAL /

TIMES

The Energy Technology Systems Analysis Program (ETSAP), IEA

www.etsap.org Source code free /Simulators to be

purchased

ENPEP-

BALANCE

Argonne National Laboratory. Energy and Power Evaluation Program

http://www.dis.anl.gov/projects/Enpepwin.html Free to Download

MESSAGE III IIASA, Laxenburg, Austria Messner S., Strubegger M., (1995), User's Guide for

MESSAGE III, IIASA, WP-95-069 Commercial

LEAP Stockholm Environment Institute – Boston

Center http://www.energycommunity.org/

Free for developing countries

IMAGE the Netherlands Environmental Assessment

Agency http://www.rivm.nl/bibliotheek/rapporten/500110002

.pdf Upon a cooperation

agreement

MERCI /

ATHDM E3 IHS, Vienna, Austria

Miess M., Schmelzer S., Balabanov T. (2010), The Austrian Hybrid Dynamic Model E3: Methodology,

Application and Validation, IHS internal WP Work in progress

Source: Authors.

PROMITHEAS-4: “Knowledge transfer and research

needs for preparing mitigation/adaptation policy portfolios”

12

MARKAL/TIMES

The acronym “MARKAL” stands for MARKet ALlocation. This model was first developed by the Brookhaven National Laboratory in the late 1970’s. In 1978 the Energy Technology and Systems Analysis Program (ETSAP) was established by the International Energy Agency (IEA) to pursue further development of the model. Since then, a whole family of models was created and most recently the TIMES (The Integrated MARKAL-EFOM4 System) model was introduced as a successor of MARKAL5.

Source: Loulou et al. (2004); Bhattacharyya and Timilsina (2010)

Specific Characteristics of MARKAL

The MARKAL model facilitates the analysis of different future energy system pathways over a

medium to long term, by integrating energy, environmental, and economic factors. Since the

development of MARKAL, many extensions were introduced. The model originally started

out using a linear programming approach that focused entirely on the integrated assessment

of energy systems. The developed amplifications went from the introduction of non-linear

programming, combining a 'bottom-up6' modelling technique with a 'top-down' macro-

economic view, to the application of stochastic programming, which allowed addressing

future uncertainties, to model multiple regions (Seebregts et al. 2001).

The model works with a user defined map (Reference Energy System) of the energy system

which contains information on the following features (Seebregts et al. 2001):

4EFOM (Energy Flow Optimization Model) is another bottom-up energy model on which the TIMES

model is based upon. 5 However, MARKAL still can be used if necessary. 6 I.e., the specific technological features of the energy sector are accounted for explicitly.



13

• Conversion of energy (e.g. power plants, refineries, solar plants)

• Primary supply of energy carriers (e.g. mining, petroleum extraction);

• Consumption of energy (e.g. industrial energy use, vehicles);

• Demand7 (exogenous, forecasts have to be produced outside the model);

• Technical characteristics;

• Technology costs.

With respect to the exogenously given end-use energy demand level, the model estimates a

discrete supply curve. Therefore, all quantities and prices are in equilibrium (suppliers

produce exactly the amount demanded by consumers) (Loulou et al. 2004). A portfolio is

provided with a cost minimising set of energy resources, energy carriers, transformation

technologies, etc; which satisfy the user defined constraints (e.g. energy balance, electrical

system operation, emission caps, technology portfolio standards, taxes, etc.). More

importantly, the model quantifies the environmental emissions resulting from this portfolio

(Seebregts et al. 2001; Johnson 2004).

The MARKAL model has typically been employed to address issues related to carbon dioxide

emission reduction, technology dynamics and R&D (Seebregts et al. 2001). “The specification

of new technologies, which are less energy- or carbon-intensive, allows the user to explore

the effects of these choices on total system costs, changes in fuel and technology mix, and

the levels of greenhouse gases and other emissions.” (Seebregts et al., 2001)

Johnson (2004) presents exemplary questions which can be investigated by the MARKAL

model:

• What happens if a new technology becomes available, or if an old one becomes

cheaper or more efficient?

• What are the implications of a technology forcing policy (e.g., a renewable portfolio

standard)?

• How do changes in technology, environmental policy, and resource availability/costs

interact?

As Bhattacharyya and Timilsina (2010) conclude, this model is among the better suited ones

to be used for the specific needs in developing countries, and in our case, emerging

economies, however, if compared to a bottom-up accounting type of model (as LEAP) it

misses some important features of these countries (e.g. including the degree of the informal

sector, energy shortages, the degree of economic transition).

7Demand can be disaggregated by sector and by functions in the sector.



14

Specific Characteristics of TIMES

The TIMES model, as it was previously stated, is based on the MARKAL modelling paradigm. The “Documentation for the TIMES Model Part I” (Loulou et al. 2004, p. 52) summarises the main differences between the MARKAL and the TIMES model. The main are:

• User defined period lengths (e.g. small steps within the first few periods,

greater durations thereafter);

• Greater user flexibility in input data specification independent of time

periods (matching is done by the model);

• User-chosen time-slices of commodities;

• Processes in different Reference Energy System sectors have the same basic

features, activated by data specification;

• Greater specification possibility of commodity-related criteria;

• Investment payments can be timed more accurately and it is possible to

define time- dependent discount rates.

The TIMES model operates with user-provided estimates of energy related equipment in all sectors, characteristics of available technologies, together with present and future energy sources and their potentials (Loulou et al. 2004, p.7).

Evaluation of MARKAL/TIMES

Urban et al. (2007; p.3478) find that MARKAL accounts for a medium number of developing countries characteristics, such as: electrification, traditional-bio fuels, urban-rural divide, subsidies, emission training and a wide assessment of renewable energies. The MARKAL or TIMES models, respectively, have been used in numerous national and regional studies. The European Commission has used the TIMES model for the evaluation of the Renewable Energy Strategy for 2020 (RES2020). Further, the TIMES model (among others) was applied to optimise the Electricity, Heat and Natural Gas Markets of the EU-25. Sulkan et al. (2010) found the MARKAL model useful in modelling alternative energy futures for Turkey. Various Estonian doctoral theses8 and studies have been conducted using MARKAL, among them “Reduction of CO2 emissions in Estonia during 2000-2030” by Agabus et al. (2007). Also in Moldova MARKAL was applied to investigate energy efficiency measures and renewable energy sources implementation possibilities. The preliminary results of this study can be found in “MARKAL Application for Analysis of Energy Efficiency measures and Renewable Energy Sources” by Robu et al. (2010) Evaluation

Criteria Description

Methodology Optimisation approach creating a dynamic partial equilibrium including all user provided energy sector specifics

8 “Long-Term Capacity Planning and Feasibility of Nuclear Power in Estonia Under Uncertain Conditions” by Landsberg (2008) and “Large-Scale Integration of Wind Energy into the Power System Considering the Uncertainty Information” by Agabus (2009).



15

Transparency, complexity, easiness of use

Technology-rich energy/economic/environmental model that requires long preparatory work

Data requirements & software requirements

Medium to high data requirement, including estimates of: energy related equipment in all sectors, characteristics of available technologies, present and future energy sources and their potentials; GAMS (General Algebraic Modeling System) is required; Windows based

Costs Cost per user for educational license: €1.200– €3.000; for the 12 beneficiaries the licensing costs could reach € 30.000

Level of coverage of M/A issues

MARKAL-MACRO: provides for endogenous and price responsive demands, and estimates of GDP impact and feedbacks; Allows certain behavioural characteristics of observed markets to be reproduced

Compliance of outputs with projects objectives

Used to simulate European Commission integrated policies on the use of renewable sources, climate change mitigation and energy efficiency improvement, the so called 20–20–20 targets, and far more stringent M/A targets in the longer term at the national and pan EU level

Availability of training and technical support

The most demanding and expensive part of MARKAL/TIMES is the training of 8 days €22.000-€30.000; however, broad documentation is available resulting from the wide use of the model; Costs of Technical support: €500-€1.800 for one year

International recognition

Most widely used bottom-up optimisation model; used in > 40 countries

Sources: http://www.etsap.org/; Loulou et al.2004; UNFCCC 2006.



16

ENPEP-BALANCE

The Energy and Power Evaluation Program (ENPEP) was developed in 1999 by the Centre for Energy, Environmental and Economic Systems Analysis (CEEESA9 - Argonne National Laboratory in the USA) and the U.S. Department of Energy (DOE). It is now used in over 80 countries. BALANCE is one of ten (10)10 integrated energy, environmental and energy analysis tools named by the UNFCCC in its 2006 report on mitigation assessment (UNFCCC 2006, p.39).

Sources: Argonne National Laboratory 2008; UNFCCC 2006 “

Specific Characteristics of ENPEP-BALANCE

The ENPEP-BALANCE Model uses the following input parameters: energy system structure, base year energy statistics (with production levels, consumption levels and prices included), energy demand growth projections as well as technical and policy constraints. With this information, an energy network is created graphically and configured by the user. The developers stress the importance of the model applying a market share algorithm. Through this it is possible to estimate the penetration of supply alternatives (Argonne National Laboratory 2008, p.1). “The equilibrium solution develops an energy system configuration that balances the conflicting demands, objectives, and market forces without optimizing across all sectors of the economy” (Argonne National Laboratory 2008, p.2).

This equilibrium solution, i.e., the set of market clearing prices and quantities, is found by the simultaneous intersection of supply and demand curves for all energy forms, as depicted in the network structure (Argonne National Laboratory 2008, p.2). Regarding environmental issues, BALANCE calculates green house gas emissions, local air pollutants (such as SOX, NOX, CO, CO2, and methane), water pollution and land use (UNFCCC 2006, p. 39; Argonne National Laboratory 2008, p.3).

9 CEEESA – Center for Energy, Environmental, and Economic Systems Analysis; 10 The nine modules are: MACRO-E, MAED, LOAD, PC-VALORAGUA, WASP, GTMax, ICARUS, IMPACTS and DAM.



17

The model allows for an annual analysis over a time horizon of up to 75 years (Connolly 2010:1069).

The model, being a simulation type of model, allows for better incorporation of non-price factors in the analysis. This is of particular importance if emerging or developing country features are being considered.

Evaluation of ENPEP-BALANCE

The model was applied in various studies11, among them the following involving participating economies of PROMITHEAS-4: A regional European project to evaluate various GHG mitigation options conducted studies for 10 countries including Bulgaria, Turkey and Ukraine; A World Bank Project to develop an Energy and Environmental Review for Bulgaria (The World Bank 2001); A CEEESA project together with the Romanian Institute of Power Studies and Design to develop a long-term energy strategy for Romania (Koritarov et al. 1998); And a CEEESA project to analyse carbon mitigation policies in Turkey conducted for the World Bank (Conzelmann et al. 2002). Furthermore, ENPEP-III (an older version of the model) is applied in Moldova.12 It is already evident from this summary that the ENPEP-BALANCE model is mostly employed to analyse national (versus regional) energy-systems (Argonne National Laboratory 2008, p.4f; Connolly et al. 2010, p.1069). Evaluation Criteria Description

Methodology Non-linear, equilibrium energy system model with economic and environmental modules; determines the response of various segments of the energy system to changes in energy prices and demand levels

Costs Can be downloaded for free from <www.dis.anl.gov/projects/Enpepwin.html>; However, training costs for 5 days amount to ~ € 7000; Costs for technical support amount to another € 7000


Medium to high: energy statistics, energy demand growth projections, technology coefficients; Windows based


The emphasis is on mitigation studies; some were already conducted in PROMITHEAS-4 participating economies


Used for green - house-gas (GHG) emissions projections and modelling the regional electricity networks; analysis of mitigation strategies


Typical training duration is 5 days for basic applications and two weeks for advanced applications; Technical support is provided via phone or email (€ 7.000 for 80 hours)


Used in over 80 countries

Sources: UNFCCC 2006, p.23, Argonne National Laboratory 2008, p. 4f; <http://www.dis.anl.gov/projects/Enpepwin.html#balance>.

MESSAGE

The acronym “MESSAGE” stands for Model for Energy Supply Strategy Alternatives

and their General Environmental Impact. The model was developed by the

11 An extensive list can be found on the developer’s web page: <http://www.dis.anl.gov/projects/Enpepwin.html#balance> 12 See “Greenhouse Gas reduction for scenarios of power sources development of the Republic of Moldova” by Robu and Comendant (2010).



18

International Institute for Applied Systems Analysis (IIASA) in the 1980s and is widely used by the International Atomic Energy Agency (IAEA) and its member states (Connolly et al. 2010:1072). The model operates similar to the MARKAL/TIMES model and the actual version is MESSAGE IV.

Sources: IIASA, Connolly et al. 2010

Specific Characteristics of MESSAGE

The user determines all the system-inherent and physical constraints and a Reference Energy System, where all the necessary configurations of the energy network are represented. Moreover, the necessary input data includes the performance characteristics of the technologies. The model then creates various energy system scenarios, which minimise total system costs, from resource extraction to the end-use. This is done starting from the base year leading up to the end of the time horizon (max. 120 years) in five to ten year steps13 (Connolly et al. 2010, p. 1072). “All thermal generation, renewable, storage/conversion, transport technologies, and costs (including SO2 and NOX costs) can be simulated by MESSAGE as well as carbon sequestration (Connolly et al. 2010, p. 1072).” Moreover, a stochastic energy system model was developed to assess key uncertainties within the energy system. This includes uncertainties concerning technological, socio-economic and climate change specifications into the modelling structure. MESSAGE was further linked with the MACRO14 model to allow for a specific treatment of the impact of policies on energy costs, GDP and on energy demand. Another important model development includes extension of the model to cover all six (6) Kyoto GHGs, their drivers and mitigation technologies. In their research project about further developments of the Kyoto-Protocol, Nakicenovic and Riahi applied the “macroeconomic model MACRO [...] to assess the economic impact and

13 http://www.iiasa.ac.at/Research/ECS/docs/models.html [Accessed 12/05/2011]; 14 “MACRO corresponds to the macroeconomic module of the top-down macroeconomic model MERGE” (Manne, Richels 1992).



19

price-induced changes of energy demand due to carbon abatement policies (Nakicenovic, Riahi 2003, p.4).”

Evaluation of MESSAGE

The report by Urban et al. (2010, p. 3478) assessing energy models for developing countries finds that many developing countries (again, similarities arise regarding IAM characteristics for emerging economies) characteristics are included. These are electrification, traditional bio-fuels, urban-rural divide, subsidies, clean development mechanism, emission trading and renewable energies. The model was applied in the following research projects: The development of global energy transition pathways for the World Energy Council (Nakicenovic, N., Riahi, K., 2001); GHG emission scenarios for the Intergovernmental Panel on Climate Change (Nakicenovic et al. 2000); Energy supply options in the Baltic states (IAEA, 2007); Moreover research projects are currently being undertaken using MESSAGE in Moldova.

Evaluation Criteria Description

Methodology Systems engineering optimisation model; Technology-rich energy systems model with economic and environmental modules


Time demanding development of case studies

Costs Free for academic purposes


Data: Energy/Economic/Environmental database, which corresponds to the EU statistical standards. Software: A free Linear Programming (LP) solver is provided. However depending on the problem complexity, more powerful LP and Non-Linear



20

Programming (NLP) solvers can be seamlessly used by the software; Windows based

Availability of the model and the Data

Economic/Energy/Environmental Data base corresponds to the EU statistical standards


The model is mostly used to estimate global or regional multi-sector mitigation strategies


With emphasis on mitigation a multitude of national studies has been completed, e.g. on options for increasing the use of renewable energy for China or energy supply options in the Baltic States, etc


The training (also conducted by IAEA) takes approximately 2 weeks; most demanding part after the initial training is the development of case studies: this can take up to half a year with IIASA team’s support


Several hundred users; wide use in IAEA member countries

Sources: Connolly et al. 2010:1072; <http://www.energycommunity.org/default.asp?action=71> [Accessed 12/05/2011].



21

LEAP

The Long-range Energy Alternatives Planning (LEAP) model was developed in 1980 in the USA. Later, the Stockholm Environment Institute (SEI) took over the maintenance and further development of the model.

Specific Characteristics of LEAP

LEAP offers broad modelling possibilities: The whole range of sectors, technologies and costs within energy-systems can be simulated. Questions regarding externalities for any pollutant, decommissioning costs and unmet demand costs can be answered. The time horizon for the evaluation of national energy-systems in LEAP typically lies between 20 and 50 years, but can be extended unlimitedly. The analysis is conducted on an annual basis (Connolly et al. 2010, p. 1071). The model requires relatively low data inputs, e.g. there is a possibility to assess energy systems and GHG emission without further information on technology costs (UNFCCC 2006, p.50). Different approaches are taken to model the demand and supply side. On the demand-side a spectrum from bottom-up, end-use accounting technique to top-down macroeconomic modelling is covered. The supply side offers a spectrum of physical energy and environmental accounting as well as simulation methodologies, which are used for developing a clear picture of the electricity power generation and for planning capacity expansions (Connolly et al. 2010, p.1071; UNFCCC 2006, p.51). The LEAP output consists of the following details: fuel demands, technology costs, unit productions, resource extraction, GHG emissions, air-pollutants, full system social-cost-benefit analysis and non-energy sector sources and sinks. “Usually, these results are then used to compare an active policy scenario versus a policy neutral business-as-usual baseline scenario (Connolly et al. 2010, p.1071; UNFCCC 2006, p.50).” Bhattacharyya and Timilsina (2010) consider this type of model to be the most suited one for addressing developing countries’ characteristics. The accounting framework makes the model very flexible regarding data requirement. However, especially the

Sources: Bhattacharyya and Timilsina 2010, Connolly et al. 2010:1071, UNFCCC 2006, p. 50f.



22

underlying scenario-based structure (versus an optimisation approach) makes them recommend this model for developing countries (Bhattacharyya and Timilsina 2010, p. 508; 513). Moreover the United Nations Framework Convention on Climate Change (UNFCCC), in its Training Handbook on Mitigation Assessment, describes the calculations of the model as non-controversial because of their simple verification and high transparency (UNFCCC 2006, p. 51).

Evaluation of LEAP

LEAP was also reviewed by Urban et al. (2007, p. 3478). They found that it includes a large number of developing countries’ characteristics (therefore, also characteristics of emerging economies), such as performance of the power sector, electrification, traditional bio-fuels, urban-rural divide, subsidies, individual assumptions per country, emission trading, clean development mechanism, renewable energies and rural energy programmes. The applications of the LEAP model are numerous (Community for Energy, Environment and Development - URL). Recently, an assessment of CCS (Carbon capture and storage) potential was conducted in Greece, to analyse the emission mitigation strategies for 2050 (Bellona Foundation 2011). Moldova used LEAP for preparing the “Second National Communication of the Republic of Moldova to UNFCCC”15. In Estonia, two studies were prepared recently on the basis of LEAP: “Energy Planning Models Analysis and Their Adaptability for Estonian Energy Sector” by Dementjeva and “Analysis of current Estonian energy situation and adaptability of LEAP model for Estonian energy sector” by Dementjeva and Siirde. Another notable study was prepared by the SEI, analysing how Europe can show leadership in keeping global climate change under the limit of 2°C higher warming.


Methodology Accounting type; Optimisation model was released on May 7th Transparency, complexity, easiness of use

Notable for its flexibility, transparency and user-friendliness

Costs Free to qualified users, but there is a cost for OECD (Organisation for

15 This can be downloaded at: <http://unfccc.int/essential_background/library/items/3599.php?rec=j&priref=7159&suchen=n >



23

Economic Co-operation and Development) based users; Paid license for EU27/free for developing countries; Whether costs arise for emerging economies has yet to be clarified. Total cost arising for the project € 8.800

Data requirement & software requirement

Data: Provides national "starter" data sets; Includes a built-in Technology and Environmental Database (TED) for a variety of technologies Software: Windows


For adaptation macroeconomic indicators (price, GDP, etc.)


Final and useful energy demand analyses; Stock-turnover for transport; Scenarios of energy and non-energy sector emissions and sinks


Online training is available but not sufficient; One week of trainer-led training is recommended; Technical support provided against fee;


Currently LEAP has over 5000 users in 169 countries

Source: Connolly D. et al. 2010, UNFCCC “Module 5.1 – Mitigation Methods and Tools in the Energy Sector” 2006:50ff.



24

IMAGE

The Integrated Model to Assess the Global Environment (IMAGE) was developed in the late 1980s16 by the National Institute for Public Health and the Environment (RIVM) in the Netherlands and is currently maintained by the Netherlands Environmental Assessment Agency (MNP). The latest version, incorporating many enhancements and extensions, created in close co-operation with different institutions in this area, is IMAGE 2.4.

Sources: Bouwman et al. 2006, p.5, Urban et al. 2007, p. 3479.

Specific Characteristics of IMAGE

The IMAGE 2.4 model is one of the most complex modelling frameworks developed until now. Different independent models can be combined for various purposes: For example, the TIMER hybrid model investigates the energy supply and energy demand side of the economies and the FAIR model analyses policy options (PBL – Netherlands Environmental Assessment Agency - URL). Human activities in areas such as industry, housing, transport, agriculture or forestry have various implications on human and natural systems. In IMAGE these interactions are specified and explored thoroughly. The “key-drivers” of the model are defined as change, population and macro economy (Bouwman et al. 2006, p. 8). The specific features of the IMAGE 2.4 framework can be divided into three (3) interacting categories (Bouwman et al. 2006, p.13):

- Socio-economic system: This category includes demographics, energy supply and

demand, agricultural demand and trade, as well as the broad category “world

economy”.

- Earth system: This category contains an explicit land use and land cover model,

including the carbon, nitrogen and water cycle, as well as the atmosphere and ocean

systems17

.

- Impacts: This category offers options for evaluating climate policies, using the policy

decision-support model FAIR18

. Hence, climate impacts, land degradation issues,

water stress, biodiversity, as well as water & air pollution can be addressed.

16 It was called Integrated Model to Assess the Greenhouse Effect. 17 An additional model, GLOBIO 3, can be used to address biodiversity issues.

Main facts



25

Evaluation of IMAGE

Urban F. et al. (2007), in their assessment of developing countries’ features that are accounted for energy models, described the IMAGE model as incorporating a medium number of developing countries’ characteristics, namely electrification, traditional bio fuels, urban-rural divide, clean development mechanism, emission trading and a wide assessment of renewable energies (Urban et al. 2007; p.3478). Research conducted on the basis of the IMAGE model includes: IPCC Special Report on Emissions Scenarios (SERS; IPCC Special Report on Emissions Scenarios, 2000), EUruralis study focusing on future prospects for agriculture and the rural areas of the EU-25 countries19, and various Greenhouse Gas Reduction studies20. Evaluation Criteria Description

Methodology Hybrid simulation model Transparency, complexity, easiness of use

Comprehensive Integrated Assessment Model (IAM) consisting of variety of sub-modules

Costs The model is useable only in close cooperation with the IMAGE developers, costs are therefore not specified;


IMAGE cannot be provided as a "ready to use package" to others: Much of the performance of IMAGE actually comes from design of scenario assumptions and its translation into actual model input. Therefore a lot of expert knowledge is necessary.


Provides insights into the full range of adaptation and mitigation options, including the costs, benefits and risks of different climate futures, policies and socio-economic development pathways, etc.


Applied in assessing climate mitigation strategies


The developers are open for serious collaborations with other institutes, to share model results, to work together on projects or model development.

International recognition Multitude of studies analysing scenarios of global and regional environmental change.

Sources: Bouwman et al. 2006; http://themasites.pbl.nl/en/themasites/image/overview/index.html.

MERCI

The Model for Evaluating Regional Climate change Impacts (MERCI) was developed by the Institute for Advanced Studies (IHS) Vienna in 2009, and is still being refined and advanced during its use in diverse projects. Since then it has been used in applied research for different Austrian ministries. MERCI is a multisectoral dynamic hybrid top-down bottom-up model, currently implemented at a national level. The main strength of MERCI lies in its ability to simultaneously depict overall economic circumstances, as well as such concerning the energy sector at a detailed technological level. Currently MERCI can be used on a national level, or for an entire region with a

18 “FAIR is widely used to assess the environmental and abatement cost implications of international regimes for the differentiation of future emission reductions of greenhouse gases. The model links long-term climate targets and global reduction objectives with regional emission allowances and abatement costs, accounting for the Kyoto Mechanisms.” Bouwman et al. 2006, p.16 19 Initiated in 2004; http://www.eururalis.eu/ 20 E.g.:http://www.pbl.nl/en/publications/2000/Global-and-Regional-Greenhouse-Gas-Emissions-Scenarios [accessed 14/05/2011]; Studies mentioned in: Bouwman et al. 2006, An extensive list can be found here: http://themasites.pbl.nl/en/themasites/image/publications/articles/index.html [accessed 14/05/2011]



26

homogenous economic structure, due to its multisectoral composition and the precondition of regional equality in prices and the state of technology.

Source: Authors

Specific Characteristics of MERCI

MERCI makes use of the hybrid, top-down, bottom-up modelling approach suggested by Böhringer and Rutherford (2008). The top-down part of the model consists of (currently 13) cost minimizing production sectors, an infinitely lived representative agent, who maximizes total lifetime utility, i.e. a composite of consumption and leisure, a government agent in charge of various political instruments such as taxes, subsidies and quotas, and an artificial agent representing foreign trade. All production and utility functions are in the form of Constant Elasticity of Substitution (CES) functions. The theoretical underlying is the classical structure of the small open economy Ramsey model. Within the bottom-up part of the model, the electricity sector is split up in currently eight (8) different technologies, all producing the same consumption good, electricity. These technologies require different input structures of labour, capital and other intermediate input goods for production, which determine their different production costs. Energy demand is taken from the top-down equilibrium, and, subject to resource and capacity constraints of each technology (e.g. locations for hydro power plants, plant capacities of processing raw energy, etc.), the most cost efficient technology mix is found within the bottom-up solution process. The top-down and the bottom-up parts of the model are solved simultaneously, generating a set of activity levels (i.e. output quantities) of production sectors and technologies, and prices of all goods and factors, such that demand meets supply in all markets. MERCI is designed to assess different possible future developments in a complex economic and ecological sense, and to evaluate them with respect to the criteria important to the user. Based on an equilibrium data set in the base year, and a calibrated long term reference path (Business As Usual, BAU), a shock is imposed on the economy, and a new equilibrium path is computed. These shocks, or scenarios, typically include unforeseen changes in the economic structure, or political instruments, in order to design economy and environment. The range is broad, and can be adapted according to nearly any focus of interest. Currently imposed scenarios are amongst others changes or introductions of taxes on



27

fossil fuels, subsidies or quotas for renewable energy sources, or a price raise in raw energy commodities.

The results (sectoral output, consumer prices, wage rates, energy mix, etc.) of the newly computed equilibrium path are compared with the BAU reference scenario. Due to the general equilibrium structure of the model, results are always in the context of the overall economy, in the form of a new equilibrium, so that important economic interdependencies are never left out. In this way the scenario effects can be analysed from different angles, with respect to several kinds of evaluation criteria.

Evaluation of MERCI



The model is formulated as a mixed complementarity problem (MCP) within the programming surrounding GAMS. Currently there is no graphical user interface, so detailed GAMS knowledge and mathematical skills are required for use.

Cost Training costs for MERCI would be free for this project.


The model is available at a development stage and needs comprehensive national level Input/Output tables and technological data

Level of coverage of M/A issues The top-down part is analyzing the adaptation options of the overall economy, while the bottom-up part depicts the technological processes on the energy level.


Could provide us national trends in the interdependency between macroeconomic issues and mitigation and adaptation strategies.


Model development is still in progress; training personnel and technical support are currently only partly available.

International recognition MERCI was used in two studies for Austrian ministries at a national level.

Source: Authors.

Conclusion

In short, the following advantages and disadvantages of the six models surveyed in this deliverable are summarised:

• MARKAL/TIMES is widely used in the research community and comes with extensive

documentation. However, it requires high computer skills and other models may be

better suited for use in emerging countries.

• ENPEP-BALANCE offers a graphical interface and was already employed in various

studies, including countries within the Black Sea Region. However, neither the study

by Urban et al. (2007) nor the study by Bhattacharyya and Timilsina (2010) analyse

ENPEP-BALANCE for its developing countries’ features. Therefore no conclusions can

be drawn regarding this issue.

• One of the drawbacks of MESSAGE is the lengthy preparation of scenarios. However,

Urban et al. (2007) find MESSAGE to be the most suited model, together with LEAP,

among the ones analysed that addresses developing countries’ specifics.

• Urban et al. (2007) and Bhattacharyya & Timilsina (2010) find that LEAP is the most

suited model available to address issues related to developing countries. Also the

low costs and the broad user-base are notable advantages. Still, until recently, LEAP

did not incorporate an optimisation tool. The actual version provides this feature.

However, this optimization module is still a work in progress, as has been noted by



28

the developers, and should be handled with reservation. However, it has already

become quite clear that LEAP is the most suited model for the mitigation/adaptation

analysis to be conducted in the PROMITHEAS-4 project.

• IMAGE 2.4’s main disadvantage regarding the focus of the PROMITHEAS-4 project is

that it is not possible to provide it as a ‘ready-to-use’ software. However, it is

capable of incorporating a medium number of developing countries’ characteristics

into the scenario analysis.

• MERCI is a complex modelling tool, whith high-level theoretical background,

flexibility of use and hybrid structure allow for a comprehensive cost-benefit analysis

when it comes to environmental and energy questions within the economy.

However, it has only been set up for Austria. Consequently, the process of

transferring the database has not been standardized yet. This may bring some

unexpected problems with it. Furthermore, MERCI has no graphical user interface

that can be easily explained to trainees, and, being a highly complex modelling tool,

would thus probably prolong the training process. Another point is that a

computable general equilibrium model, often imposing rigorous assumptions on the

economy, might not be best suited for use in emerging economies.

The next deliverable (D 2.2) will evaluate the models presented, also taking into account the conclusion from this report that the LEAP model is the most suited one for the analysis, using the following criteria, which were decided upon in the ad-hoc working group’s protocol – “Chain of activities for concluding with policy portfolios” – 4th of March, 2011, Athens:

a. The choice will be restricted to models used at European level;

b. The wideness of the model in covering mitigation/adaptation issues (The model that

is closer in covering these issues will be taken into consideration);

c. Transparency, complexity and easiness in using the model;

d. Availability of inputs (available in statistics books, national accounts);

e. Flexibility of the model in building scenarios (e.g.: a simulation model does not

impose bias in modelling outputs);

f. Compliance of outputs with our contractual obligations (socio-economic,

technological penetration);

g. Cost of acquiring the model;

h. International recognition of the model (used by governments);

i. Training and technical support.



29

References

Agabus H. (2009). Large-Scale Integration of Wind Energy into the Power System

Considering the Uncertainty information. Tallinn. TTÜ Press.

Agabus H., Landsberg M., Tammoja H. (2007). Reduction of CO2 emissions in Estonia

during 2000-2030. Oil Shale. Vol. 24, No. 2 special.

Argonne National Laboratory (2008). “ENPEP Brief Description” [Can be downloaded at

<http://www.dis.anl.gov/projects/Enpepwin.html#balance>]

Balabanov T., Miess M., Schmelzer S. (2010). The Austrian Hybrid Dynamic Model E3: Methodology, Application and Validation, IHS WP-10-12 Bayer J. L. (2009). The EU Solidarity Fund, ADAM Policy Workshop, Brussels, May 14, 2009 Bellona Foundation (2011). A Bridge to a Greener Greece, [Can be downloaded at <http://cdn.globalccsinstitute.com/sites/default/files/fil_BELLONAprintFINAL.pdf>] Böhringer C., Rutherford T.F. (2008). Combining bottom-up and top-down, Energy Economics volume 30, March 2008: 574-596 Bouwman A. F., Kram T., Goldewijk K. (2006). Integrated modelling of global environmental change – An overview of IMAGE 2.4 Comendant I., Robu S. (2010). Gas reduction for scenarios of power sources development of the Republic of Moldova. Roblems of Regional Energetics. No. 1(12). [Source: http://ieasm.webart.md/archive_ro/] Connolly D. et al. (2010). A review of computer tools for analysing the integration of renewable energy into various energy systems. Conzelmann G., Koritarov V. (2002). Turkey energy and environmental review. Argonne National Laboratory; <http://www.dis.anl.gov/news/TurkeyUndp.html> Dementjeva N. (2009). Energy Planning models Analysis and Their Adaptability for Estonian Energy Sector. Tallinn, TTÜ Press Dementjeva N. and Siirde A. (2010). Analysis of current Estonian energy situation and adaptability of LEAP model for Estonian energy sector. Energetika. 2010. T.56.No.1.pp.75-84 Dickinson T., (2007). The Compendium of Adaptation Models for Climate Change: First Edition, Adaptation and Impacts Research Division, Environment Canada, 52 pgs. European Commission (2009). White Paper - Adapting to climate change: Towards a European framework for action, COM(2009) 147 final Gavras P. (2010).The Current State of Economic Development in the Black Sea Region, Policy report I, Commission on the Black Sea Heaps Ch., Erickson P., Kartha S., Kemp-Benedict E., (2009). Europe’s Share of the Climate Challenge: Domestic Actions and International Obligations to Protect the Planet, Stockholm Environment Institute Hulme M., Neufeldt H., Colyer H. (eds.) (2009). Adaptation and Mitigation Strategies: Supporting European Climate Policy. The Final Report from the ADAM



30

Project. 2009 Tyndall Centre for Climate Change Research, University of East Anglia, Norwich, UK International Atomic Energy Agency (IAEA) (2007). Analyses of energy supply options and security of energy supply in the Baltic states, <http://www.iaea.org/OurWork/ST/NE/Pess/assets/TE_1541_ balticstudyFeb07.pdf>. Intergovernmental Panel on Climate Change (IPCC) (2007). Fourth Assessment Report (AR4), Climate Change: Impacts, Adaptation and Vulnerability, Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson, Eds., Cambridge University Press, Cambridge, UK, 976 pp IPCC Special Report on Emissions Scenarios, 2000 <http://www.ipcc.ch/pdf/special-reports/spm/sres-en.pdf> accessed 16/05/2011 IPCC Third Assessment Report (AR3) - Climate Change 2001 (2001). <http://www.grida.no/publications/other/ipcc_tar/> accessed 13/05/2011 Johnson T. (2004). An EPA Overview: Energy Technology Assessment and Regional MARKAL Modeling Initiatives” [Source: www.nrel.gov/analysis/seminar/docs/t_johnson_may04_nreltalk.ppt, accessed 10/05/2011] Koritarov V.S., et al. (1998). Development of a Fuel Policy: Energy Supply and Demand Study for Romania. Prepared for The World Bank by Argonne National Laboratory in association with the Institute of Power Studies and Design. Final Report. Volumes 1-3. Argonne National Laboratory. Argonne, IL, USA. August Kram T. et al (editor) (2006). Integrated modeling of global environmental change: An overview of IMAGE 2.4, Netherlands Environmental Assessment Agency (MNP), Bilthoven, October 2006 Landsberg M. (2008). Long-term Capacity Planning and Feasibility of Nuclear Power in Estonia Under Uncertain Conditions. Tallinn, TTÜ Press. Lavalle, C. (2009). Mainstreaming adaptation to extreme events. Deliverable D-A2.5b - Task A2.5 of the ADAM (Adaptation and Mitigation Strategies: Supporting European Climate Policy) project Loulou R., Goldstein G., Noble K. (2004). Documentation for the MARKAL Family of Models

Manne, A. and R. Richels, (1992). Buying Greenhouse Insurance: The Economic Costs of

CO2 Emissions Limits, The MIT Press, Cambridge, MA, USA.

Messner S., Strubegger M., User's Guide for MESSAGE III, IIASA, WP-95-069, 1995 <http://www.iiasa.ac.at/Admin/PUB/Documents/WP-95-069.pdf> Miess M., Schmelzer S., Balabanov T. (2010). The Austrian Hybrid Dynamic Model E3: Methodology, Application and Validation, IHS internal WP Nakicenovic N., Riahi K., (2001). An assessment of technological change across selected energy scenarios, Energy Technologies for the Twenty-First Century, World Energy Council (WEC), London, UK (available from WEC). Reprinted as RP-02-005. International Institute for Applied Systems Analysis, Laxenburg, Austria. Nakicenovic N., Riahi K., (2003). Model Runs with MESSAGE in the Context of the Further Developments of the Kyoto-Protocol. WBGU Special Assessment Report. IIASA, Laxenburg, Austria. Nakicenovic N., Alcamo J., Davis G., de Vries B., Fenhann J., Gaffin S., Gregory K., Gruebler A., et al. (2000). Special Report on Emissions Scenarios, Working Group III of the Intergovernmental Panel on Climate Change, IPCC, Cambridge University



31

Press, Cambridge, UK, 595 pp. (ISBN 0-521-80493-0) <http://www.grida.no/publications/other/ipcc_sr/> Parson E.A., (1995). Integrated assessment and environmental policy-making: in pursuit of usefulness. Energy Policy 23:463–75 Parson E.A., and Fisher-Vanden K., (1997). Integrated Assessment Models of Global Climate Change, Annual Review of Energy and the Environment, Vol. 22: 589-628 Robu S., Bikova E., Siakkis Ph., Giannakidis G. (2010). MARKAL Application for Analysis of Energy Efficiency in Economic Activities of the Republic of Moldova and Feasible use of Renewable Energy Sources. Problems of the Regional Energetics. No.2(13). [Source:<http://ieasm.webart.md/archive_ro/>] Rutherford, T. F. (1995). Constant Elasticity of Substitution Functions: Some Hints and Useful Formulae, manuscript, University of Colorado Seebregts A. J. et al (2001). Energy/environmental Modeling with the MARKAL Family of Models The Pan European TIMES model for RES2020 (2009). Model description and definitions of Scenarios <www.res2020.eu> The United Nations Framework Convention on Climate Change (UNFCCC) (2006). Module 5.1 – Mitigation Methods and Tools in the Energy Sector [Source: <http://unfccc.int/resource/cd_roms/na1/mitigation/index.htm>, accessed: 11/05/2011] The United Nations (1996). Republic of Bulgaria: the first national communication on climate change, <http://unfccc.int/resource/docs/natc/bulnc1.pdf> Urban F., Benders R. M. J., Moll H. C. (2007). Modelling energy systems for developing countries, Energy Policy Van Beeck, N. (1999). Classification of Energy Models. Tilburg University, Tilburg, The Netherlands. Van Beeck, N. (2003). A new method for local energy planning in developing countries. Tilburg University, Tilburg, The Netherlands. Ph.D. Thesis. Weyant J., Davidson O., Dowlatabadi H., Edmonds J., Grubb M., et al. (1996). Integrated assessment of climate change: an overview and comparison of approaches and results, In Climate Change 1995: Economic and Social Dimensions of Climate Change, ed. JP Bruce, H Lee, EF Haites. Contrib. Work. Group III to 2nd Assess. Rep. IPCC. Cambridge, UK: Cambridge Univ. Press The World Bank (2001). Bulgaria: Energy-Environment Review. A World Bank Country Study. Washington DC. June URLs

Community for Energy, Environment and Development <http://www.energycommunity.org/default.asp?action=71> [accessed 12/05/2011] PBL – Netherlands Environmental Assessment Agency <http://themasites.pbl.nl/en/themasites/image/overview/index.html> IIASA – International Institute for Applied Systems Analysis <http://www.iiasa.ac.at/Research/ENE/model/stochastic.html> [accessed: 17/05/2011]



32

PPRROOMMIITTHHEEAASS –– 44

Task Leader: Ptof. Bernhard Felderer

Institute of Advanced Studies (IHS),

Vienna, August 2011

““SSeelleeccttiioonn ooff MMooddeellss ffoorr

MMiittiiggaattiioonn//AAddaappttaattiioonn PPoolliiccyy””



33

This report has been read and commented by all members of the PROMITHEAS-4 Scientific Committee.

It was also disseminated for comments, through BSEC – PERMIS and BSEC – BC, to all relevant governmental and business authorities and partners before its finalization.

Partners from the beneficiary countries* of the consortium were encouraged to contact direct national authorities, agencies, institutions and market stakeholder for comments before the finalization of this report (Annex 1).

List of PROMITHEAS – 4, Scientific Committee:

18. Prof. Dimitrios MAVRAKIS, NKUA – KEPA (GREECE) -Editor 19. Dr. Popi KONIDARI, NKUA – KEPA (GREECE) – Assistant to the editor 20. Dr. Harry KAMBEZIDIS, NOA (GREECE) 21. Prof. Bernhard FELDERER, IHS (AUSTRIA) 22. Prof. Bilgin HILMIOGLU, TUBITAK – MAM (TURKEY) 23. Prof. Vahan SARGSYAN, SRIE – ESC (ARMENIA) 24. Prof. Dejan IVEZIC, UB – FMG (SERBIA) 25. Prof. Mihail CHIORSAK, IPE ASM (MOLDOVA) 26. Prof. Agis PAPADOPOULOS, AUT – LHTEE (GREECE) 27. Prof. Alexander ILYINSKY, FA (RUSSIA) 28. Prof. Anca POPESCU, ISPE (ROMANIA) 29. Prof. Andonaq LAMANI, PUT (ALBANIA) 30. Prof. Elmira RAMAZANOVA, GPOGC (AZERBAIJAN) 31. Dr. Lulin RADULOV, BSREC (BULGARIA) 32. Prof. Arthur PRAKHOVNIK, ESEMI (UKRAINE) 33. Prof. Sergey INYUTIN, SRC KAZHIMINVEST (KAZAKHSTAN) 34. Prof. Alvina REIHAN, TUT (ESTONIA)

*Turkey, Armenia, Serbia, Moldova, Russia, Romania, Albania, Azerbaijan, Bulgaria,

Ukraine, Kazakhstan, Estonia.

The EU, the Consortium of PROMITHEAS – 4 and the members of the Scientific Committee do not undertake

any responsibility for copyrights of any kind of material used by the Task Leaders in their report. The

responsibility is fully and exclusively of the Task Leader and the his/her Institution.

Acknowledgments: The Task Leader of this report acknowledges the contribution of Mr. Michael-Gregor Miess and Mr. Stefan Schmelzer for the development of this overview.



34

Table of contents

Table of Abbreviations 35

Introduction 37

Introduction 37

Mitigation and Adaptation 39

ENPEP-BALANCE 41

MARKAL/TIMES 45

MERCI 51

LEAP 54

IMAGE 58

MESSAGE 62

Conclusion 65

Definition of Terms 68

References 69

URLs: 72

Communication 72



35

Table of Abbreviations

ADAM Title of the “Adaptation and Mitigation Strategies: Supporting European Climate Policy” project Title of the “Adaptation and Mitigation Strategies: Supporting European Climate Policy” project AOS Atmosphere-Ocean System CEEESA Centre for Energy, Environmental and Economic Systems Analysis

CGE Computable General Equilibrium

CO2 Carbon Dioxide COMMEN

D Community for Energy, Environment and Development

EC European Commission

EFOM Energy Flow Optimization Model

EIS Energy-Industry System

ENPEP Energy and Power Evaluation program

ETSAP Energy technology and Systems Analysis Program

FAIR Framework to Assess International Regimes for differentiation of commitments

GAMS General Algebraic Modeling System

GDP Gross Domestic Product

GHG GreenHouse Gas

GNU GNU’s not Unix

GWP Global Warming Potential

IAEA International Atomic Energy Agency

IAM Integrated Assessment Model

IHS Institut für Höhere Studien (Institute for Advanced Studies)

IIASA International Institute for Applied Systems Analysis

IEA International Energy Agency

IER Institut für Energiewirtschaft und Rationelle Energieanwendung

IMAGE Integrated Model to Assess the Global Environment

INPRO International Project on Innovative Nuclear Reactors and Fuel Cycles

IPCC Intergovernmental Panel on Climate Change

IPE Institute of Power Engineering

LEAP Long-range Energy Alternatives Planning

M/A Mitigation/Adaptation

MAGICC Model to Assess Greenhouse-gas Induced Climate Change

MARKAL Market Allocation (Model)

MERCI Model for Evaluating Regional Climate change Impacts

MESSAGE Model for Energy Supply Strategy Alternatives and their General Environmental (Impact)

MNP Milieu en Natuur Planbureau (Netherlands Environmental Assessment Agency)

NAPA National Adaptation Programs for Action

NEEDS New Energy Externalities Developments for Sustainability

NPV Net Present Value

OECD Organisation for Economic Co-operation and Development

PBL Planbureau voor de Leefomgeving (Netherlands Environmental Assessment PET Pan European TIMES model

Pg C Petagrams (1015 g) of Carbon

PPP Purchasing Power Parity

R&D Research and Development

RES Reference Energy System/Scenario



36

RES Renewable Energy Sources

RIVM Rijksinstituut voor Volksgezondheid en Milieu (National Institute for Health and Environment)

SAM Social Accounting Matrix

SEI Stockholm Environment Institute

TES Terrestrial Environment System

TIMES The Integrated MARKAL-EFOM System

TIMER The Regionalized Energy Model of IMAGE 2.4

UNEP United Nations Environment Programme

UNFCCC United Nations Framework Convention on Climate Change

USAID United States Agency for International Development

VEDA VErsatile Data Analyst



37

Introduction

The purpose of this report is to present an in-depth classification of the previously presented models (ENPEP-BALANCE, MARKAL/TIMES, MESSAGE, MERCI, LEAP and IMAGE; see work package 2.1) according to the following criteria:

j. The choice will be restricted to models used at European level;

k. The wideness of the model in covering mitigation/adaptation issues (The model that is closer

in covering these issues will be taken into consideration);

l. Transparency, complexity and easiness in using the model;

m. Availability of inputs (available in statistics books, national accounts);

n. Flexibility of the model in building scenarios (e.g.: a simulation model does not impose bias in

modelling outputs);

o. Compliance of outputs with our contractual obligations (socio-economic, technological

penetration);

p. Cost of acquiring the model;

q. International recognition of the model (used by governments);

r. Training and technical support.

Considering them in turns: Since the Black Sea area and its economies are largely influenced by the European Community, it is reasonable to use models that are recommended and appreciated within these countries (analysed under criterion a. Except for MESSAGE, where detailed information for applications at the European level is lacking, and MERCI, which only has been used on a national scale (Austria) so far, all models were already used within the European Community. This is not surprising, since all of the considered models are very well known within the academic community. Also the international recognition (criterion h.) of the tools plays an important role, when deciding what model to apply in our analysis. Most models are known world-wide, especially MARKAL, LEAP and ENPEP-BALANCE. The second criterion to be investigated is the ability of the model to cover mitigation and adaptation issues. Regarding the models analysed here, Patt et al. 2009 (p. 385) conclude that “process-oriented models [e.g. IMAGE and MARKAL] with considerable physical detail” are less suited for dealing with adaptation issues. Simpler models calculating mitigation costs and climate damages at an aggregate scale are regarded to be more useful here. After reviewing the relevant literature, the following statement can be made: All tools can and were used for mitigation analysis already. However, the exact conduct of modelling adaptation measures is harder to assess, since documentations make no reference to the relevant issues. Regarding the point ‘Transparency, complexity and easiness in using the model’ (c. criterion): The models under consideration here are all bottom-up models, some of which include or can include top-down aspects (hybrid models). They differ in the modelling approach: optimisation or accounting. The bottom-up structure enables the models to provide a “detailed technological



38

representation of the energy system and can be used to analyse the environmental effects as

well” (Bhattacharyya and Timilsina 2010, p. 512). All tools considered herein have transparent structures, and most of them provide a graphical user interface (MERCI is an exception; however, a graphical user interface is being constructed). More complex models, requiring expert knowledge, include: MARKAL/TIMES, MERCI and IMAGE. MESSAGE can be described as an intermediate model, regarding its complexity. On the contrary, ENPEP-BALANCE and especially LEAP are convenient to use and highly transparent. The next criterion, namely the availability of inputs for model runs, (criterion d.), is a relevant issue, especially for the PROMITHEAS-4 investigation, as some details may be unattainable. This section therefore presents the relevant data inputs of each model. The optimisation models are usually more data intensive, i.e., require substantial information on demand and supply levels. Moreover, base year statistics have to include a wide range of energy specifics. Accounting type, simulation models do not need as much information, and the data requirements are determined by the type of analysis the user is about to conduct (see LEAP). It is further important for the assessment of the quality of the set of models to determine how flexible they are in building scenarios, (criterion e.). The easiness with which models can develop new scenarios follows from the model structure that will be presented in detail at the beginning of each model’s section. Nowadays, due to the rising demand for scenario analysis, this feature has been incorporated into many IAMs (Integrated Assessment Models). Especially flexible tools for scenario analysis are LEAP and IMAGE, noting that both apply a simulation modelling approach. Analysing the compliance of outputs with our contractual obligations, (criterion f.), it can be observed that the set of models was already chosen such that socio-economic aspects can be analysed with considerable detail. However, only IMAGE incorporates a detailed feed-back mechanism, mapping the effects of emissions back to the earth system. An important aspect of our considerations is the suitability of models with respect to emerging economies characteristics. This analysis is based on the description of Urban et al.

(2007) and Bhattaacharyya and Timilsina (2010). Regarding the financial aspects of the models, (criterion g.), there is great variability between them. Also the training and support costs differ substantially among the models. First, an overview of mitigation and adaptation specific issues arising with integrated assessment models is presented. Then, the report is structured as follows: Each section starts with a short review of the model in question. Each model is then analysed subject to the different criteria. The paper concludes with a summary and the selection of the most suited model, according to our contractual obligations.



39

Mitigation and Adaptation

The focus of the PROMITHES-4 project is the analysis of mitigation and adaptation possibilities within the Black Sea region’s countries, Kazakhstan and Estonia. Therefore, the model that is about to be chosen for the analysis has to be suited for these requirements. This section shall depict in more detail the difficulties that arise for such an analysis, especially with the requirement to model adaptation measures. Mitigation, in the context of strategies for the reduction of climate change related damage, was defined as follows by the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC): Mitigation consists of “[t]echnological change and substitution that

reduce resource inputs and emissions per unit of output. Although several social, economic

and technological policies would produce an emission reduction, with respect to climate

change, mitigation means implementing policies to reduce GHG emissions and enhance

sinks” (IPCC 2007, Annex I, p.818). Further, the IPCC defined adaptation as “[i]nitiatives and measures to reduce the

vulnerability of natural and human systems against actual or expected climate change effects.

Various types of adaptation exist, e.g. anticipatory and reactive, private and public, and

autonomous and planned. Examples are raising river or coastal dikes, the substitution of

more temperature shock resistant plants for sensitive ones, etc” (IPCC 2007, Annex I, p. 809). It consists of adjustments in ecological, social, or economic systems to a new or changing environment. Adaptation measures therefore seek to reduce harm or exploit beneficial opportunities in the context of climate change (IPCC 2001, Glossary, p. 708). The UNFCCC points out that mitigation21 can be investigated by two (2) different modelling approaches: top-down and bottom-up. The top-down approach is better suited for broader macroeconomic and fiscal policies (e.g. carbon taxes), while the bottom-up approach (ENPEP-BALANCE, MARKAL/TIMES, MESSAGE) ensures a specific sectoral and technology-based perspective (UNFCCC 2006, p. 10). However, this is no strict dichotomy. Many models started to incorporate both aspects. These are called hybrid models (MERCI, MARKAL-MACRO, ENPEP, IMAGE). Moreover, there exists a different category of models, namely accounting type models (LEAP), which have yet another approach towards integrated assessment modelling. In assessing mitigation strategies, it is important to analyse the implementation of mitigation measures in the three (3) energy end-use sectors, which are commercial/residential/institutional buildings, transportation and industry. Together with the energy-supply side of the economy, agriculture, forestry and waste management sectors, these sectors constitute the necessary structure for an appropriate analysis of policy scenarios (IPCC 1996, p. 3). An important issue arises with the appropriation of climate-related benefits of mitigation and adaptation. These benefits accrue at different geographic levels. Benefits from mitigation actions can be enjoyed at a global scale, while adaptation related benefits arise mostly at an individual, organisational or local level. This implies that mitigation involves many top-down 21 Of course, many other issues can be addressed with these different modelling approaches. However, mitigation and adaptation are the focal points of this study.



40

characteristics, while adaptation can be investigated better via a bottom-up approach (Patt et

al. 2009, p. 390). Looking at the structure of integrated assessment models it is challenging to account for adaptation policies appropriately. IAMs are used on a broad range of assessments of climate change related mitigation scenarios; difficulties arise, however, when it comes to adaptation. Various approaches have been considered, with different and, mostly, unsatisfying results. The range covered by IAMs goes from including no adaptation at all, modelling it implicitly and, recently, to account explicitly for adaptation. First, in modelling the damage from climate change, IAMs included estimates of the amount of adaptation that was likely to be undertaken. Second, effort was directed towards implicit assumptions about the amount of adaptation necessary to minimise climate change damages. However, this approach, known as the ‘Ricardian analysis’, was criticised because of its sole focus on partial equilibrium22 analysis and for its missing representation of frictional costs, when shifting from one production system to another23. Third, lately models have begun to include adaptation explicitly. This is done by formulating adaptation in terms of a control variable. These models, however, have shortcomings too, as there has been little progress in “including

adaptation as a more nuanced variable” (Patt et al. 2009, p. 385ff; last citation from p. 388).

22 „[I]t does not consider changes in prices of different commodities as the entire production shifts.“ (Patt et al. 2009, p.387) 23 The Ricardian approach ignores these shifts between production systems (Patt et al. 2009, p.387).



41

ENPEP-BALANCE

The Energy and Power Evaluation Program (ENPEP), used in conjecture with BALANCE, is one of ten24 available integrated energy, environmental and economic analysis tools. It is a bottom-up simulation, iterative equilibrium model, applicable locally, regionally and globally. It imitates consumer and producer behaviour, considering different constraints and signals (UNFCCC 2006, p. 13, 39). The analyst using ENPEP-BALANCE builds a representation of the region’s (or the nation’s) energy system in the software’s graphical user interface. This makes the model accessible without having to acquire extensive syntax training. Within the energy network, each sector is modelled separately, consisting of a variety of nodes connected by links between them. The node types available are depicted in Figure 1 below:

Figure 1. [Source: CEEESA, p. 7]

The range of nodes reaches from energy supply, via economic and resource processes, as well as conversion mechanisms, to energy demand. These nodes are submodels described by sets of quantity and price equations. This decentralized decision making process allows for optimal energy choices according to the decision makers’ own needs (CEEESA I, p.5ff, 12; CEEESA 2008, p. 1). Moreover, “[t]he model employs a market share algorithm to estimate the penetration of

supply alternatives”(CEEESA 2008, p.1). It can account for various competing fuels and technologies at decision nodes. Each market share reacts to changes in prices relative to the prices of alternative commodities. Together with the market share algorithm, an additional lag factor allows for the possibility of delays in capital stock turnover. These features lead to the

24 The nine tools are: MACRO-E, MAED, LOAD, PC-VALORAGUA, WASP, GTMax, ICARUS, IMPACTS and DAM.



42

nonlinear, iterative equilibrium solution, subject to various policy constraints, of ENPEP-BALANCE (CEEESA p. 12; CEEESA 2008, p. 2). After the specification of all constraints and adding the available information, ENPEP-BALANCE finds market clearing prices. This equilibrium balances demand and supply curves for the whole energy network (CEEESA I p.2, UNFCCC 2006, p. 13).

a. Use of the model at European level

A whole list of ENPEP-BALANCE applications in Europe can be found on CEEESA’s homepage (CEEESA I25). The wide use of the model in assessing future options for a sensible treatment of climate change related issues in eastern and south-eastern Europe is striking. Among them are eight countries for which mitigation and adaptation policy portfolios shall be developed within the Promitheas-4 project: Albania26, Armenia27, Bulgaria28, Moldova29, Romania30, Russia31, Turkey32 and Ukraine33. Moreover, ENPEP-BALANCE was employed in two projects involving Greece34. These studies analysed GHG (Green House Gas) emission projections and mitigation scenarios, investigated the cost of meeting EU environmental standards (on Turkey’s fossil-fired power plants), conducted energy and nuclear power planning scenarios, energy and environmental reviews and characterised old and inefficient technologies (CEEESA I).

b. The wideness of the model in covering mitigation/adaptation issues

As was mentioned previously under a.), ENPEP-BALANCE was applied in various studies investigating mitigation issues. “Numerous countries used ENPEP to help prepare GHG mitigation assessments as part of their national communications to the UNFCCC” (UNFCCC 2006, p. 46). As was described in the introduction above, it is of great importance for an integrated assessment model regarding mitigation to describe the energy end-use and supply side sectors in detail. Herein lies ENPEP-BALANCE’s strength, since the user defines the energy network most appropriate for the region or nation in focus. ENPEP-BALANCE enables the calculation of environmental residuals associated with a pre-defined energy system configuration. Air pollutants, such as GHGs, sulphur oxide (SOx), nitrogen oxide (NOx), carbon monoxide (CO) and carbon dioxide (CO2), methane (CH4), volatile organic compounds (VOC), leads (Pb) and others can be considered within these

25 http://www.dis.anl.gov/news/EnpepwinAppsEurope.html 26 Project title: “Capacity Building in GHG Mitigation Analysis for Balkan Countries”; 27 Project title: “Energy and Nuclear Power Planning Study for the Period up to 2020”; also used within the IAEA TC Project (2000-2003); 28 Project titles: “Bulgaria Energy and Environmental Review (EER)”, “Bulgaria GHG Emission Projections”, “Bulgaria UNFCCC National

Communications”, “Infrastructure Development and Nuclear Competitiveness”, “Capacity Building in Energy and Power Systems Analysis in Bulgaria”; 29 Project titles: “Moldova UNFCCC First National Communication”, “UNDP Project "Climate Change Enabling Activity (Phase II)" Technology Needs Assessment”; See report “Overview of models in use for Mitigation / Adaptation policy”, for details. 30 Project title: “Capacity Building in Energy and Power Systems Analysis in Romania”, “Developing a Fuel Policy for Romania”, “Romania UNFCCC National Communication”; 31 Project title: “Modeling of Heat Sources in Power System Expansion Planning”; 32 Project title: “Finding the Most Cost-Effective Sulfur Control Strategy for Turkey's Yatagan Lignite-Fired Power Plant”, “Capacity

Building in Energy and Environmental Systems Analysis in Turkey”, “Analyzing Turkey's GHG Mitigation Options”, “Costs of Meeting EU Environmental Standards; on Turkey's Fossil-Fired Power Plants”, “Infrastructure Development and Nuclear Competitiveness”, “Providing Modeling Support for Turkey's First National Communication to the UNFCCC”; 33 Project title: “Modeling and Analysis of GHG Emissions in Ukraine: Selecting and Adapting the ENPEP Program to Ukrainian Conditions and Test Modeling”; 34 Project titles: “Integrated Resource Planning for the Island of Crete”; “Capacity Building in GHG Mitigation Analysis for Balkan Countries”;



43

residuals. Noteworthy, however, is the possibility to analyse the effect of water pollution, waste generation and land use via these environmental residuals (CEEESA 2008, p. 3). This provides a particularly informative model when calculating the costs and benefits of planned mitigation measures.

c. Transparency, complexity and easiness in using the model

From the description of the model, it can be concluded right away that ENPEP-BALANCE is a transparently structured tool. The model’s graphical user interface adds to the easiness in using the model. The user, constructing a RES (Reference Energy System), can build a graphical map of the energy system, where various drop-down menus assist the configuration of the mentioned nodes and links. Some complexity arises, however, from large data requirements. This makes building new scenarios somehow more time consuming.

d. Availability of inputs

The required data inputs consist of information on the energy system structure, especially base year energy statistics. These have to include production and consumption levels, as well as prices. Moreover, it is important to provide estimates of future energy demand growth exogenously. Additional constraints on technology and policy issues are of course further decision parameters for the users (CEEESA 2008, p. 1). Characterising behavioural relationships, which is necessary in simulation models, can be challenging as knowledge of these parameters is lacking, and even more so in countries where time series data is missing (UNFCCC 2006, p. 13).

e. Flexibility of the model in building scenarios

The ENPEP-BALANCE model is well suited for scenario building. This feature is documented in the various reports35 from mitigation studies (many of which regard GHG emissions or carbon mitigation). Moreover, since ENPEP-BALANCE is a simulation model, it is easier to account for non-price factors in the analysis compared to optimisation models (UNFCCC 2006, p. 13).

f. Compliance of outputs with our contractual obligations

Although the UNFCCC clearly recommends using this model, neither Bhattacharyya and Timilsina nor Urban et al. make any reference to ENPEP-BALANCE in the context of developing countries issues. From the list of applications, it can be concluded, however, that this model is especially useful and flexible for usage in developing countries, as well as in developed countries. (See h.) As was previously mentioned, the model incorporates a market share algorithm to calculate the penetration of energy technologies. This percentage market share of a supply option is sensitive to the price of the commodity relative to the prices of other commodities. Furthermore, the market share algorithm is sensitive to user defined constraints, government policies, consumer preferences and the “ability of markets to respond to price signals over time” (CEEESA 2008, p. 1f). However, no feedback mechanism from the resulting emission level from the production and consumption side of the economy is incorporated.

g. Cost of acquiring the model

The model can be downloaded free of charge from: <www.dis.anl.gov/projects/Enpepwin.html>

h. International recognition of the model

35 An extensive list is provided in CEEESA 2008, p.4f;



44

ENPEP-BALANCE found use in over 50 countries, where it was employed, among many other institutions36, by energy and environmental ministries. Although the list of applications of ENPEP-BALANCE is long and applications can be found all over the world, in North/South America, Africa, Europe as well as in Asia, some exemplary studies are mentioned below: In the United States, CEEESA conducted a study analysing carbon emission mitigation strategies. Venezuela, Jordan and Kazakhstan investigated their GHG mitigation options as part of their national communication to the UNFCCC. Further, Egypt constructed an energy plan of the transport sector considering especially environmental issues (CEEESA II).

i. Training and technical support

The model developers recommend at least five days of training. The associated costs amount to around 7000 €. Technical support is offered by phone, e-mail or online. Basic support is provided for free, an extensive premium support package can be acquired for approximately 7000 € (UNFCCC 2006, p. 47).

36 Electric utilities, power merchants, transmission companies, consulting companies, lending agencies and research institutions. (CEEESA 2008, p. 4)



45

MARKAL/TIMES

The MARKet ALlocation (MARKAL) model is currently maintained by the Energy Technology and Systems Analysis Program (ETSAP), which was established by the International Energy Agency (IEA) in 1978. A whole family of models evolved is based on MARKAL and the most prominent model version used today is the TIMES (The Integrated MARKAL-EFOM37 System) model, which was developed in 1999 and is expected to replace MARKAL over time. The MARKAL modelling technique can be summarised as follows: The bottom-up model employs an optimisation method (MARKAL: linear programming; MARKAL-MACRO: non-linear programming; MARKAL with uncertainties: Stochastic programming), i.e., it develops options for energy supply services which minimise the total cost of the energy supply system. This minimisation is undertaken considering user imposed constraints, such as limits on technology, CO2, etc., and for an exogenously specified demand. For each technology either the utility maximising (MARKAL-MACRO) or the producer/consumer surplus maximising (MARKAL/TIMES) prices and quantities are calculated over the entire planning horizon. Through the optimisation approach, the model is an especially useful tool where many technical aspects have to be studied and the future development of costs is well known (UNFCCC 2006, p. 12, 26; Bhattacharyya and Timilsina 2010, p. 512). Before continuing with a description of TIMES, a few details about EFOM, the Energy Flow Optimization Model, will complete the analysis. Developed in the 1970s by Finon at the ‘Institut Economique et Juridique de l’Energie’ at Grenoble, France, EFOM became a prominent model around the world. In short, it is a multi-period system optimisation model. It uses linear programming as the solution mechanism to minimise total discounted system costs, constrained on an exogenously given energy demand level. Each sector can be either investigated separately (“single-sector mode of analysis”) or, using a different configuration, the whole energy system can be analysed (“multi-sector model”). Together with the fact that the electricity industry is extensively covered by the model, this makes it especially useful for the analysis in developing countries (Bhattacharyya and Timilsina 2010, p. 511). TIMES, the latest development within this family of models, merges features of both, the MARKAL and the EFOM model. The analyst can either decide to calculate the least-cost solution for the whole system or can focus on a specific sector. Moreover, investment and operating decisions can be included in the analysis. On the one hand, demand drivers’ characteristics and elasticities of demand (regarding the demand drivers and prices), that are exogenously specified, facilitate a sensible demand-side analysis. On the other hand, a supply-side analysis is conducted using a set of supply curves for the spectrum of available resources. As elaborated above, the TIMES model maximises producer and consumer surplus. This yields a partial equilibrium solution. Summarising, it results that TIMES is a more flexible replacement for MARKAL (Bhattacharyya and Timilsina 2010, p. 512f, Seebergts et al.

2001).

37 EFOM - Energy Flow Optimization Model



46

Figure 2. [Source: ETSAP II]

As depicted in Figure 2 the MARKAL and the TIMES model generators can be used more or less equivalently. However, as was previously outlaid in the model overview (W.P. 2.1) and above, there are some important technical differences in TIMES:

• User specified, completely flexible time period lengths;

• Data is decoupled from the initial period, i.e., “user provides technical and cost data at those

past years when the investment actually took place, and the model takes care of calculating

how much capacity remains in the various modelling periods.” Changing the initial period or

the period length is therefore much easier than in MARKAL;

• User chooses time slices for every commodity/process (seasonal/monthly, weekly, daily);

• Processes in all RES (Reference Energy Systems) have the same basic features, which are

activated via data specification;

• Completely flexible Processes;

• Investment and dismantling lead-times costs;

• Vintage processes and age-dependent parameters;

• Commodity related parameters (e.g. total production, total consumption, flow variables);

This way, the user imposes limits and costs on commodities;

• More accurate and realistic depiction of investment cost payments;

• The concentration of CO2, radiative forcing and global temperature change (stemming from

GHG emissions) is endogenised through a set of variables and equations.

(Loulou et al. 2005, p. 52ff) a. Use of the model at European level

The TIMES model was used in the PET model (Pan European TIMES model), to enable the analysis of the renewable energy targets set by the European Union for 2020. This is a technically oriented model which illustrates in detail the whole energy system of the EU-27 member states38 for the period from 2000 to 2050. This project characterised future options

38 Iceland, Norway and Switzerland were also included in the study.



47

for policies and measures. Moreover, specific targets for the renewable energy sources’ contribution were calculated that can be achieved by implementation of these options. Additionally, the implications of the achievement of these targets to the European economy was investigated (RES2020). MARKAL/TIMES tools are used in the USAID project “E&E Regional Energy Security and Market Development” which is conducted by Armenia, Georgia, Moldova and Ukraine.39 Further, the TIMES model generator was employed within the Pan European TIMES model of the New Energy Externalities Developments for Sustainability (NEEDS) project. “The

ultimate objective of the NEEDS Integrated Project [was] to evaluate the full costs and benefits (i.e.

direct + external) of energy policies and of future energy systems, both at the level of individual

countries and for the enlarged EU as a whole” (NEEDS). Another noteworthy use of the TIMES model at EU level is the ‘EU30 TIMES-Electricity and Gas supply model’, which seeks to optimise the electricity, heat and natural gas markets of the EU member states. Studies using this model are mostly conducted at the ‘Institut für Energiewirtschaft und Rationelle Energieanwendung’ (IER) at Stuttgart University, Germany (IEA/ETSAP 2008, p. 58). European countries using MARKAL/TIMES include: Belgium, Finland, France, Germany, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom (IEA/ETSAP 2008, p. 77 - 177).


MARKAL/TIMES can support the following list of mitigation measures: transportation, energy demand, energy conversion and supply, energy sector emissions, non-energy sector industrial process emissions, solid waste management, geological sequestration and the value of carbon rights. Moreover, MARKAL-MACRO can inform about the effect of macroeconomic policies, such as carbon taxes or emission caps (UNFCCC 2006, p. 29). Additionally, TIMES gives the analyst a powerful tool in addressing climate issues through a set of variables and equations that endogenise the concentration of CO2, calculate radiative forcing and global temperatures change (from GHG emissions and accumulation) (Loulou et

al. 2005, p. 55) . c. Transparency, complexity and easiness in using the model

A key advantage of MARKAL/TIMES is its transparency: “Data assumptions are open and each result may be traced to its technological roots” (Johnson 2004, p. 10). The user defines a Reference Energy System (RES), wherein all energy sources, conversion processes and end-use possibilities are included. However, the MARKAL/TIMES model generators require high skilled users or analysts.40 On the one hand, the high data requirements add some complexity to the usage of the model. On the other hand, the lengthy training and the sophisticated programming approach hamper the easiness of the model, although they have advantages for different types of analyses.


39 http://www.winrock.org/fact/facts.asp?CC=5830&bu= 40 See criterion i.)



48

As indicated by UNFCCC (2006, p. 12, 23) MARKAL and the hybrid sister-model MARKAL-MACRO are quite data intensive (medium-high data requirement). This necessitates a data and results handling system, as can be seen in Figure 2. The two databases that can be linked to the model are ANSWER and VEDA-FE/VEDA-BE. These two systems are quite different substitutes. ANSWER is more user friendly, but this comes at the cost of pre-defined output tables. Contrary to this, VEDA-BE gives the user full flexibility in structuring and exploring the model’s results. It is recommended to use VEDA-BE for the TIMES model, that uses ANSWER as a data handling system (IEA/ETSAP II).


As previously mentioned under criterion c., scenarios are based upon RES. This, together with the modelling approach of optimisation, implies some bias in the model’s output. However, the developers managed to make energy demand price-responsive, so that a more realistic tax policy analysis or an investigation about the effects of emission constraints is possible. Furthermore, TIMES can account for multiple periods and various regions, as well as explore uncertainties connected to future energy system development trajectories (Bhattacharyya and Timilsina 2010, p. 512f).


MARKAL/TIMES model generators use a bottom-up optimisation approach towards their least-cost solution. Bhattacharyya and Timilsina (2010, p. 501), conclude that, although bottom-up models are better suited for developing countries characteristics than top-down models, optimising models are less useful than accounting-type models. Neither MARKAL, nor TIMES can include informal sector characteristics, or such important features (for the context of developing/emerging countries) as energy shortages or subsidies (possible, but typically not included). However, they do account for rural energy features, rural-urban divide and non-price policies. Moreover, TIMES can include economic transition variables (Bhattacharyya and Timilsina 2010, p. 503). The MARKAL model constructs optimal future scenarios by either optimising over all time periods, i.e., under the assumption of perfect foresight, or year-on-year, i.e., using myopic expectations (UNFCCC 2006, p. 12).41 The time horizon is user controlled. Typically, the development of a scenario is analysed over a period of 20-50, sometimes 100, years (Connolly et al. 2010, p. 1072). MARKAL/TIMES quantifies the sources of emissions from the associated energy system and calculates estimates of energy and material prices, demand activity, technology and fuel mixes, the marginal value of individual technologies to the energy system, GHG and other emission levels, as well as mitigation and control costs (Johnson 2004, p. 11, UNFCCC 2006, p. 28). According to Mr. Goldstein, from the Department of Energy in the United States and expert on MARKAL/TIMES models, the level of CO2 emissions is calculated and reported in the units, in which the user has specified the emissions. Further, MARKAL/TIMES computes total costs of the energy system. Apart from this, no extra social costs are calculated (unless this is reflected in the cost data fed into the model). Moreover, the costs resulting from mitigation/adaptation scenarios can be calculated for different target groups, such as industry, agriculture, households, government, etc., assuming that data is provided for the relevant sectors. The model can further calculate the percentage of the penetration of renewable energy sources or the penetration of energy efficient technologies as far as these technologies are

41 Agents having myopic expectations act short-sighted, i.e., they do not have perfect foresight over the entire model horizon.



49

specified by the user. Finally, the model only calculates administrative costs, if the cost data is integrated into the model.


The source code for the MARKAL/TIMES model generators is distributed for free42. Since the code is written using the commercial software package GAMS (General Algebraic Modeling System), this has to be purchased. Moreover, a data handling system (ANSWER or VEDA-FE) and a solver (MINOS, CPLEX, XPRESS, GUROBI or CONOPT) has to be acquired. This leads to total costs of around 1.275 €-3.170 € for an educational license and around 9.825 € - 15.200 € for a commercial license (Connolly et al. 2010, p. 1072; IEA/ETSAP I).


MARKAL is currently used in around 70 countries by 250 institutions. In the category of optimisation models, it is probably the one most widely used and definitely the best known (Connolly et al. 2010, p. 1071, Seebregts et al. 2001, Bhattacharyya and Timilsina 2010, p. 512). Figure 3 below gives a clear picture of the huge acknowledgment MARKAL/TIMES have received. Moreover, the Final report of Annex X (2005-2008) “Global Energy Systems and Common Analyses” summarises the studies and projects using MARKAL/TIMES as modelling tools. ‘The IEA Energy Technology Perspective Project’ and ‘The ETSAP TIMES Integrated Assessment Model’ used Global (TIAM). These are/were two prominent global applications (IEA/ETSAP 2008, p. 21ff; 26ff). At national level, the MARKAL-family models were employed in the following country studies: MARKAL-MACRO was used to conduct a model for Kazakhstan, a TIMES model was developed for the Russian Federation, the MARKAL model was used for energy modelling in Portugal (together with LEAP), Spain investigated future energy policies under the European Energy and Climate Policy Framework, Norway modelled a GHG emission reduction of 75% until 2050, France calculated CO2 emissions reduction using MARKAL/TIMES, etc (IEA/ETSAP 2008, p. 77 - 177).


Training in using this model is the quite demanding, as it can take some months (Connolly et

al. 2010, p. 1072). The costs for an eight (8) days training workshop vary between 20.000 € and 30.000 €, and the support costs amount to 350-1.800 € per year (UNFCCC 2006, p. 24).

42 A letter of Agreement has to be signed.

PROMITHEAS-4: “Knowledge transfer and research needs for preparing mitigation/adaptation policy

portfolios”

50

Figure 3. [Source: IEA/ETSAP 2008, p. ii]



51

MERCI

The Model for Evaluating Regional Climate change Impacts is a dynamic hybrid top-down bottom-up Computable General Equilibrium (CGE) model. The combined model structure is based on Böhringer and Rutherford (2008). The Institute for Advanced Studies in Vienna, (IHS), has adapted this hybrid modelling structure for developing MERCI since 2009 (Balabanov et al., 2010). There is also a static version of the model, focussing primarily on issues concerning labour market analysis.


MERCI is a very young model, and therefore is currently still only applicable at a national level for Austria. Extensions for neighbouring countries or the European level are currently being implemented. MERCI has been used for 2 national evaluations for Austrian ministries in the last year: In a study for the Federal Ministry of Labour, Social Affairs and Consumer Protection, the potential of “green jobs” in the near future, as well as general impacts of energy and environmental policies on labour market issues were analysed (Balabanov et al. 2010). Most recently, in a study for the Federal Ministry for Economy, Family and Youth, an analysis of the overall economic impacts of investment incentives created through public subsidies in specific sectors of the economy, as well as the impact of subsidies on the actual investment is being conducted with the help of the static version of MERCI (Miess et al. 2011).


MERCI is a scenario based general equilibrium model. A “business as usual” scenario is being compared to a set of scenarios dealing e.g. with different mitigation and adaptation strategies. Various scenario tools like subsidies for sustainable energy resources, green quotas, carbon taxes or other emission taxes, emission caps and control limits, exogenous shocks on resource prices, etc., are included in the modelling framework. The sectoral structure allows for a detailed assessment of emissions not only of the energy sector and its different technologies, but also for industrial process emissions within the intra-sectoral chain of intermediate consumption. Transportation, detailed depiction of industrial production, sectoral and household energy consumption demand, energy conversion and supply, are some of the mitigation/adaptation measures that are incorporated in the model, and make MERCI a powerful tool for mitigation and adaptation analysis on a regional level. The variety and flexibility in using and defining different scenario variables not only allow MERCI to optimize and create a dynamic adjustment path in order to reach a predefined future goal, it is also possible to set certain scenario variables to specific predefined values, and “see what happens” until the end of the model horizon. While using the model in any of these analysing procedures, the total economic and social abatement costs are calculated automatically (in the form of GDP, welfare, sectoral growth, emissions, etc.), due to the general equilibrium framework used in MERCI.


MERCI does not have any kind of user interface. Therefore, it is necessary for the potential user not only to be familiar with the modelling software GAMS, but also to have at least basic



52

knowledge in General Equilibrium theory, and an understanding of Mixed Complementarity Problems in order to use the model with already incorporated scenarios. Building new scenarios requires even more knowledge of the aforementioned technical issues, and generally requires a serious amount of preparation/training time for each user. Therefore, all analyses conducted with the model until now were carried out by IHS personnel only. Until a proper user interface is being developed, distribution of the model to other institutions would only be appropriate if general equilibrium modellers (and a corresponding amount of their time) are available at the receiving institution.


The main data structure used in MERCI is that of standard input/output tables, which are released by most of the countries in the world nowadays. A slight preparation of these datasets in the form of a Social Accounting Matrix (SAM43) is required, which takes some weeks’ time, but is not hard to conduct, given that the data needed for the procedure are available. Energy statistics in the base year, concerning all available technologies for energy production, are also necessary. This data is implemented at the technological level in a similar way as the I/O tables on the sectoral level. The second type of data required in MERCI, are elasticities. The range is broad as in any Computable General Equilibrium (CGE) model, covering price elasticities in production, consumption elasticities of households, elasticities between leisure and consumption, etc. These elasticities may vary strongly between countries, and not all of them might be available in all countries of the Black Sea region, Kazakhstan and Estonia. Estimating missing elasticities can be a hard and expensive task, adapting elasticities from countries with similar structures can be dangerous and misleading when analysing results, due to wrong assumptions on habits and economic structures.


As previously described under b), MERCI is a very powerful tool when developing scenarios. Political, environmental, or economic scenarios can be implemented with various different scenario tools. The structure of the model, and the way of using it, by directly changing source code whenever trying a new model run, allows implementing whatever scenario instrument the modeller can literally imagine. This being the greatest freedom, it is also the greatest disadvantage, because developing new scenarios usually needs expertise in economic theory, strong mathematical background and some programming skills.


Reductions of GHG emissions are measured in tCO2 equivalents, and can be displayed for each sector in the economy, each electricity or energy technology, or end user. The display of marginal abatement cost curves is optional within MERCI, depending on the focus of pre-specified scenario definitions. Social costs can be derived from various model outputs, like an explicit measurement of general welfare, GHG emissions, and additional tax burdens, or even financial ease (e.g. subsidies) for the specific household agents in the model, who represent the whole population. However there is not an aggregate measure for social costs in Euro. Due to the sectoral model structure costs for different target groups as industry, services, agriculture and other economic sectors, as well as different household groups, and the government sector (administrative costs) can be displayed in high detail. Still, while such implementation costs of specified scenarios can be measured in monetary units (Euro), effects of GHG Emissions, or temperature changes can not.

43For more information on the SAM, please see e.g. King (1985), Pyatt (1988), or Reinert and Roland-Holst (1997)



53

A main focus of MERCI lies on the detailed depiction of the electricity and energy sector. Due to that, the percentage of RES penetration or the penetration of energy efficient technologies, can be delineated more than sufficiently.


Licenses for the most recent GAMS distribution and for the PATH solver (which can be purchased online) are required to use the model properly. A single user license is available for $640 for academic and $3,200 for commercial use. Also multiple user licenses are available (GAMS-SALES). IHS would provide the source code of MERCI for free for this project. Yet the training time that would be required would definitely exceed the person months allocated to IHS.


Recently MERCI was used in 2 national studies for the Austrian government, as already described in criterion a. (Balabanov et al. 2010), (Miess et al. 2011).


Model development for MERCI is still in progress. Training personnel and technical support of IHS are currently only partly available. However there is a huge community of modellers, (teachers and students alike) available on the internet, which provides technical support for particular problems encountered when using or developing CGE models implemented in GAMS. (GAMS-L)



54

LEAP

The Long-range Energy Alternatives Planning model (LEAP) was developed in 1980. The Stockholm Environment Institute (SEI) is currently in charge of the model. Noteworthy is the wide use of this model: its user-base counts over 5000 users in 169 countries (Connolly et al. 2010). LEAP’s technique proceeds at two steps: First, the straightforward and consistent energy, emissions and cost-benefit accounting problems are solved by a calculator-like tool. At the second step, the analyst specifies further aspects of the model he/she wants to analyse including tabular formulas. These then help to define time-varying data or multi-variable models. Based on this, econometric analysis and simulation can be conducted (Heaps 2008, p.6). Due to the described methodological environment, the LEAP model ends up being a powerful analytical tool. The accounting framework enables simple, transparent and intuitive investigation of the complex matter of energy policies (UNFCCC 2006, p.15; Heaps 2008, p.5). This year (2011) a new version of LEAP was published: Optimization modelling is now supported. The additional feature was developed in collaboration with various institutions44 and is based on the Open Source Energy Modeling System (OSeMOSYS). This again evolved from the GNU45 Linear Programming Kit (GLPK)46. Although, both software packages are open source, they were incorporated into LEAP. Therefore, LEAP’s optimisation extension is well integrated into the stand-alone modelling tool (Heaps 2011, p. 1; SEI 2011, p. 4). Why use this optimisation tool47? “Typically you will use this new capability to calculate the

optimal expansion and dispatch of power plants for an electric system, where optimal is defined as the

energy system with the lowest total net present value of the social costs of system over the entire

period of calculation (from the base year through the end year).” (Heaps 2011, p. 1) a. Use of the model at European level

Among the high quantity of LEAP users are various European countries. As previously mentioned in the report “Overview of models in use for Mitigation / Adaptation policy”, researchers in Estonia, Moldova, Albania and Greece have already carried out studies and research using LEAP. The most recently published report is “A Bridge to a Greener Greece” (Bellona Foundation 2011). Furthermore, Heaps et al. examined the role Europe can play in keeping global climate change targets in a study conducted at SEI “Europe’s Share of the Climate Challenge - Domestic Actions and International Obligations to Protect the Planet” (Heaps et al. 2009).


44 The Stockholm Environment Institute worked together with the International Atomic Agency (IAEA), the United Nations Industrial Development Organization (UNIDO), the UK Energy Research Center and the Royal Technical University (KTH) in Sweden. 45 GNU stands for “GNU’s not Unix” 46 “[A] software toolkit intended for solving large scale linear programming problems by means of the revised simplex method.” (Heaps 2011, p. 1) 47 The optimisation extension in LEAP2011 is not yet fully finished and “should be used for testing purposes only.” (Heaps 2011, p. 1)



55

As an accounting model LEAP is able to consider mitigation issues in a highly flexible and realistic way. In contrast to an optimization based model, it does not impose a clear bias in modelling scenarios. Various studies with the focus on investigating mitigation issues have already been based on LEAP. Therein, a comparison between a business-as-usual scenario and one or more mitigation policy scenario(s) was prepared. These analyses outline how aggressive mitigation measures to reduce energy demand affect the energy composition of a country or region and the corresponding emission levels affect the economy. LEAP can compute a preliminary mitigation assessment using essential production-based emission variables and elaborated production statistics, i.e., ‘Tier 1’ emission factors, provided by the IPCC, which are included in LEAP. The analysis becomes more precise as more data on local and regional emission sources and levels are provided. Great emphasis is put on the provision of data regarding chemical compositions of the used fuels, since this specifies the emission estimates (Davis 2010a and 2010b, SEI 2006, p. 4). As with adaptation, LEAP does not assess the cost of it. There is, however, the possibility to include externality values, which are associated with different GHGs and local air pollutants. In turn, these externality costs may be integrated into the overall net present value (NPV) calculation.

Figure 4. [Source: COMMEND (II)]


LEAP is a highly transparent and flexible tool. It is an accounting type model, i.e., its modelling approach is based on the quantitative representation of flows of energy. These are defined through simple engineering relations. The demand for energy is modelled through various methods: Either a bottom-up, an end-use accounting or a top-down macroeconomic



56

approach is applied. To model electric sector generations and calculate capacity expansion on the supply side in a satisfying way, accounting and simulation methods are employed. Moreover, these methods allow for the inclusion of results from other models, which investigate these supply side factors in a more thorough way (UNFCCC 2006, p. 15; Heaps 2008, p. 6). Transparency for the user is guaranteed by the accessible user interface, which is designed as depicted in Figure 4. This is the ‘Analysis View’, where the data structure provided by the analyst is organized in the hierarchical tree on the left-hand side and the data for a specific section of the energy system (here: household demand) can be viewed in the window on the right-hand side. The great advantage of this tool can be seen in the third window: immediately a chart can be drawn to provide a solid graphical interpretation of the data (Heaps 2008, p. 8). LEAP incorporates even more tools for data and result presentation. Additionally to charts, also tables and maps can be developed right away (and can even be exported to Microsoft Excel, Word and PowerPoint).


The data requirements of a LEAP based model will crucially depend on the type of analysis the researcher wants to investigate. Therefore, no exhaustive list of necessary data can be provided. Furthermore, since some/many elements48 of LEAP can be added or removed by the user, and while the user decides whether he/she is about to conduct a top-down or bottom-up analysis, these decisions will shape the structure of the required data set (SEI 2006, p. 1). Next, it is important to make sure that the investigation is not determined by the available data. Instead, the detail necessary for our study of mitigation/adaptation issues would have to be clearly defined and data would have to be provided accordingly. The analyst therefore, does not conduct his/her study according to the available data, but seeks the necessary data for his/her study. A full overview of data requirements is given in SEI (2006). Data concerning demographics, the economic situation, the energy system, demand characteristics, transformation possibilities, the environmental variables and fuel specifics can be incorporated within the analysis.


As already described under b.), analysts using LEAP can build various scenarios: First, they will develop a baseline scenario, which will then be used for comparison. Second, they create scenarios with differing levels of mitigation and other measures. These are long-term scenarios, accounting for the socio-economic and environmental impacts of the possibly undertaken policies. It is very convenient to build scenarios using LEAP. The model software provides a Scenario Manager, where policies can be investigated for their individual effects or for their composite effects (Heaps 2008, p. 6).


It was previously mentioned in the report “Overview of models in use for Mitigation / Adaptation policy”, how useful LEAP is in covering specific characteristics of developing countries. Reviewing them shortly, LEAP can include and evaluate variables such as electrification, traditional bio-fuels, urban-rural divide, individual assumptions per country, emission trading, renewable energies, rural energy programmes. These can be accounted for explicitly. Moreover, implicitly modelled characteristics are the performance of the power sector and subsidies (Urban et al. 2007, p. 3478).

48 Such as transformation analysis, pollution and GHG emissions analysis, cost analysis, and non-energy sector GHG accounting.



57

Users of LEAP can deduct various cost-specific outcomes from the cost-benefit summary report provided as one part of the output. Therein, a summary of costs and benefits for each computed scenario relative to the pre-specified baseline (or “business as usual”) scenario is delivered.49 The cost summary compares “total cumulative emissions of all greenhouse gases avoided by each scenario (shown in terms of the global warming potential of those pollutants in tonnes of

CO2 equivalent)” (SEI 2011, p. 22, emphasis by authors). Furthermore, LEAP can calculate the individual and combined Global Warming Potential (GWP) of one or more greenhouse gases. For this calculation, the user can choose among two measurement units: Carbon (C) equivalents and Carbon Dioxide (CO2) equivalents. According to Mr. Heaps, developer of LEAP, social cost for the different scenarios can be calculated in any currency. Moreover, they can be presented in discounted or in real terms. LEAP calculates the total social net present values of a computed scenario against a pre-specified baseline scenario. However, it does not specify the actual costs accruing to different groups, i.e. it deals with costs rather than prices. However, the NPV can be split up to determine the streams from each sector (households, industry, transport, services, electric supply, etc.). LEAP can calculate the percentage of RES penetration, as well as the penetration of energy efficient technologies. Finally, administrative costs can be included optionally, provided that the necessary data is included.


LEAP is free for analysts in developing countries. However, users from OECD (Organisation for Economic Co-operation and Development) countries are charged a fee for the use of the model (Connolly et al. 2010, p. 1071).


This model is widely recognized among government agencies, academics, non-governmental organizations, consulting companies and energy utilities. Together with the mentioned magnitude of users (over 5000) this makes LEAP a standard model for integrated resource planning and greenhouse gas mitigation assessment (Heaps 2008, p. 5).


Support for any occurring problems with LEAP is easy to access by phone, email or a web forum. It is provided for free to registered users. COMMEND – Community for Energy, Environment and Development and the respective web site <www.energycommunity.org> is a rich source on various LEAP and energy related issues, especially for developing countries’ analysts. Moroever, the community itself provides trainings and support (UNFCCC 2006, p. 66f). Although LEAP is an intuitive tool, some training is necessary nevertheless. A minimum of five days is recommended and offered for free, however, paying for the expenses (travel costs, etc. for the trainer).

49 The user can freely change the monetary unit of the summary.



58

IMAGE

IMAGE, the Integrated Model to Assess the Global Environment, was developed in the late 1980s at the National Institute for Public Health and the Environment (RIVM), Netherlands. Today, a separate institution, the Netherlands Environmental Assessment Agency (MNP) is maintaining and further developing the model. After a series of revisions, the latest version on the way to IMAGE 3 is 2.4 (MNP 2006, p. 5). IMAGE can be described as a hybrid model, incorporating both the top-down and the bottom-up views of energy and economic systems. The main focus of the model is to investigate the global interconnectedness of the natural/biophysical and the human/socio-economic systems. The targeted issue of IMAGE is therefore a very broad one, particularly different from the other models included in this study. The model investigates “direct and indirect pressures on human and natural systems closely related

to human activities in industry, housing, transport, agriculture and forestry” (MNP 2006, p. 8). On the one hand, the socio-economic conditions and interactions are analysed focusing on 24-regions (plus Greenland and Antarctica).50 On the other hand, the climate, land-cover and land-use dynamics are modelled by applying a geographically explicit approach, namely a grid resolution of 0.5 by 0.5 degrees. This feature contributes to the strength of the model in the analysis of relations within the earth energy system (MNP 2006, p. 8f). The model connects three modules for the system analysis: The Energy-Industry System (EIS), the Terrestrial Environment System (TES) and the Atmosphere-Ocean System (AOS). The EIS calculates emissions for all the regions, considering industrial and energy sources. This calculation is undertaken by the TIMER51 energy model. The TES provides emission quantities related to global land-cover changes and other indicators ((agro-)economic and climate characteristics). The resulting emission levels from the calculations within the EIS and the TES are then used to simulate the greenhouse-gas stock in the atmosphere, using AOS (MNP 2006, p. 9ff). The modules themselves consist of a complex structure that is depicted in Figure 5. The socio-economic effects on land use change affect the climate, which in turn again affects the human system. Demographics, energy supply and demand levels, as well as agricultural production and economic interactions at the world level produce emissions and imply a certain land allocation. These emission levels and the land use affect the biophysical system. The resulting climate impacts, the land degradation, the degree of water pollution, the effects on biodiversity as well as the air pollution are identified. Based on these results the FAIR52 model investigates possible policy options (MNP 2006, p. 12).

50 The regions are: Canada, USA, Mexico, Central America, Brazil, Rest of South America, Northern Africa, Western Africa, Eastern Africa, Southern Africa, Western Europe, Central Europe, Turkey, Ukraine region, Kazakhstan region, Russia, Middle East, South Asia, Korea region, East Asia, Southeast Asia, Indonesia, Japan, Oceania, Greenland and Antarctica. 51 TIMER = The Targets IMage Energy Regional Model, developed in connection with the IMAGE 2.2 version. 52 FAIR = Framework to Assess International Regimes for differentiation of commitments



59

Figure 5. Source: MNP 2006, p. 13


IMAGE was applied within the EUruralis study, which was initiated in 2004. This study investigated future prospects for agriculture and the rural areas of the EU-25 countries. Moreover, the project “Greenhouse Gas Reduction Pathways in the UNFCCC Process up to 2025” employed an older version of the model, IMAGE 2.2, in the analysis. Further, IMAGE was applied in the ADAM (Adaptation and Mitigation) project, funded by the Framework 6 Programme by the European Community, in the development of two mitigation and adaptation scenarios. Finally, IMAGE was used in the ENSEMBLES project, also financed by the Framework 6 Programme, to prepare an ‘ambitious climate policy scenario’ (MNP I).


The results of the climate model IMAGE can be fed into FAIR to analyse environmental and mitigation issues. This model then calculates the implied abatement costs of different future



60

emission reduction scenarios. These costs are calculated by marginal abatement cost curves of the energy- and industry-related CO2 emissions (MNP 2006, p. 187). “The model has been developed to explore and evaluate the environmental and abatement cost

implications of various international regimes with respect to the differentiation of future commitments

for meeting long-term climate targets, such as the stabilization of atmospheric GHG concentrations” (MNP 2006, p. 187). The model computes who shall contribute which quantity in reaching climate stabilization goals. However, the analysis of these questions is not solely focused on cost and technical assessment; the emphasis is put also on responsibility and equity matters (as far as they can be assessed in a quantitative way) (MNP 2006, p. 187). The most recent studies conducted with IMAGE to assess mitigation potentials are listed at the MNP web page (MNP II). The list includes ‘Oil and natural gas prices and greenhouse gas emission mitigation’ (van Ruijven and van Vuuren 2009) and ‘Assessment of bottom-up sectoral and regional mitigation potentials. Background report’ (Hoogwijk M. et al. 2008).


This description of the model leads to the conclusion that IMAGE 2.4 is a complex model with far-reaching implications. Additionally to an intuitive user interface, the model’s structure is easy to understand. Applying this model, however, necessitates quite a wide range of specialised and committed users. Therefore, it can be used only in research projects carried out together with MNP (see g. & i.).


Under the auspices of the Netherlands Environmental Assessment Agency researchers developed a History database of the Global Environment (HYDE). This database provides historical time series (mainly for the period between 1890-2000, for some data, between 1700-2000) for land-use and land-cover data, population data, livestock, GDP, energy-specific data, production levels, estimates of energy consumption levels, as well as figures of atmosphere, oceans and terrestrial environment characteristics (MNP 2006, p. 94). Although, the database provides all this information for many countries, further and more recent information has to be provided regardless.


IMAGE is a simulation-type model and therefore produces less biased scenarios than optimisation-type models. The wide use of IMAGE in scenario development, within projects at the MNP, projects conducted by the European Community or the IPCC confirms this suitability (MNP 2006, p. 6).


Urban et al. (2007) came to the conclusion that IMAGE is a moderate tool in modelling developing countries characteristics. Explicitly traditional bio-fuels, clean development mechanism, emission trading and a broad spectrum of renewable energies can be included. Implicitly, electrification and the urban-rural divide are modelled (Urban et al. 2007, p. 3478). According to Mr. van Vuuren, Senior researcher at Netherlands Environmental Assessment Agency (PBL), IMAGE does not account for social costs of different scenarios. Neither does it compute the costs resulting from mitigation/adaptation scenarios differentiated along sectors (households, industry, production, agriculture), but the emission reduction by sector is



61

calculated naturally. The administrative costs of the scenarios are not included.53 IMAGE does, however, compute the penetration of renewable energy sources. The cost implications of different scenarios produced with IMAGE are analysed by the FAIR sub-model. The model uses marginal abatement cost curves calculated by the data (on energy- and industry-related carbon dioxide emissions) provided by the TIMER54 2.0 model and the IMAGE model (that provides data on carbon sinks, MNP 2006, p. 187). As a measure of potential economic impacts of emission reduction, the FAIR model applies the ratio of abatement costs to GDP in PPP (MNP 2006, p. 195). The calculations related to GHG emission reduction are conducted in Petagrams of Carbon (Pg C) but can be calculated in percentage terms as well.


The cost of acquiring the model is not specified, since the model cannot be provided as a ‘ready-to use’ software package. IMAGE can be used only by close cooperation with MNP and other partner institutions (MNP IV).


As mentioned under e., IMAGE was used in various studies and research projects around the world. It was a central constituent within the IPCC Special report on Emission Scenarios, IPCC Representative Concentration Pathways, the UNEP Third and Fourth Global Environment Outlook, the Millennium Ecosystem Assessment, the OECD Environmental Outlook, as well as many more (MNP III).


Close collaboration with the developers is necessary to use this model (as mentioned under g.). Further, the model’s developers point out that a high level of “expert knowledge is necessary to make good use of it” (MNP IV).

53 Transaction costs in international trading are, however, included. 54 TIMER – The Regionalized Energy Model of IMAGE 2.4



62

MESSAGE

The Model for Energy Supply Strategy Alternatives and their General Environmental Impact (MESSAGE) came into being at the International Institute for Applied System Analysis (IIASA) in Austria around 1980. It is widely used in the member countries of the International Atomic Agency (IAEA). Two major extensions were later developed: ‘The Macroeconomic Model MACRO’ and ‘The Model to assess Greenhouse-gas Induced Climate Change’ (MAGICC). Moreover, a stochastic and a myopic model exist (IIASA; Connolly et al. 2010). MESSAGE is a ‘systems engineering’ optimisation tool. It determines energy scenarios, which minimise total system costs considering user imposed constraints on the energy system. Further, MESSAGE analyses ‘how much of the available technologies and resources are actually used to satisfy a particular end-use demand [...].’ Based on this and further information provided by the analyst, an energy system configuration is developed for the entire planning horizon (base year – > user defined end of the time horizon) (IIASA II). The technological details supplied by the user will vary from geographical and temporal characteristics of the problem under consideration. MESSAGE is rather flexible with respect to the provided degree of this information. When utilizing this model generator, the analyst describes a RES, specifying all relevant links and nodes within the energy system. This RES should include all necessary performance characteristics of technologies (IIASA II).

As was previously mentioned above, two extensions of MESSAGE exist: First, ‘The Macroeconomic Model MACRO’ was developed based on the top-down macroeconomic model MERGE (Model for Evaluating the Regional and Global Effects of GHG Reduction Policies). At IIASA, MACRO was modified to achieve full compatibility with MESSAGE, and MACRO is now predominantly used in connection with MESSAGE. MACRO solves the inter-temporal utility maximisation problem of a representative producer-consumer for each region. Introducing this top-down view on the economy the effect of policies on the energy system (e.g.: energy costs, GDP, energy demand) can be investigated. “The link is established

by using MESSAGE results on total and marginal costs of energy supply to derive the quadratic

demand functions for MACRO. The linked model is iterated until MACRO’s resulting energy demands

do not deviate from MESSAGE’s by more than a given fraction” (IIASA I).

The second extension, The Model to Assess Greenhouse-gas Induced Climate Change (MAGICC), estimates aggregate climate impacts of Energy, Economic and Environment scenarios. By incorporating a carbon cycle model55 and estimating resulting emissions, implied net carbon flows and atmospheric CO2 concentrations, changes in radiative forcing, temperature and sea level relative to 1990 can be calculated (IIASA I). Rao and Riahi (2006, p. 179) used MAGICC in connection with MESSAGE to achieve scenario consistency regarding the proposed forcing target. The previously obtained emission levels from MESSAGE are provided as inputs for MAGICC. This tool then estimated the hypothetical forcing resulting from the given emission levels. The result of this calculation, the new CO2 concentration56 limit, is then returned back to MESSAGE. Consistency between the GHG

55 Describes how atmospheric inputs (emissions) and outputs (physical and chemical sink processes) affect changes in the atmospheric carbon concentration (IIASA I). 56 http://www.iaea.org/INPRO/index.html



63

emissions from MESSAGE and the forcing target is obtained through iteratively repeating this procedure (Rao and Riahi 2006, p.179f).


Our partner institution from Moldova, the Institute of Power Engineering (IPE ASM), pointed out that its personnel are being trained in using the model for further research uses. Also, partners from the Energy Strategy Center of Scientific Research Institute of Energy (SRIE-ESC) in Armenia have used the model within their studies. Moreover, local studies were conducted in Armenia using this model. MESSAGE is currently used in the International Project on Innovative Nuclear Reactors and Fuel Cycles (INPRO) founded by the IAEA. The set of members consists of all 32 IAEA member states as well as the European Commission.


MESSAGE is primary a tool for mitigation analysis: Instead of focusing on climate targets, MESSAGE estimates the effects of mitigation strategies at regional and global levels within various sectors. (Connolly et al. 2010, p. 1072) Moreover, MESSAGE was applied to “develop climate mitigation scenarios aimed at achieving long-term stabilization of global radiative forcing” (Rao and Riahi 2006, p. 177). In MESSAGE it is possible to impose emission control limits on individual plants and on one or more groups of plants. Although it is possible to model emission trading among plants or utilities, this is a more complicated issue (Rogner 2002, p. 37).


The current version, MESSAGE V, incorporates a user interface for data entries and program calls. The IAEA characterises it as “an extremely flexible model”, however the analyst has to develop the RES and clarify the relevant policy questions (Rogner 2002, p. 40). As an optimisation tool, with demanding data requirements, the scenario development can be quite challenging. The data structure and modelling technique, however, provide a transparent analysis framework.


The research necessary to provide data for MESSAGE shall obtain information on the available energy resources, future technological developments and the evolution of technological parameters over time (Messner and Strubegger 1995b, p. 11). Moreover, information on energy demand needs to be provided exogenously, together with seasonal variation in demand. An advantage of MESSAGE is the absolute flexibility in energy and fuel demand specification for the user (Rogner 2002, p. 21).


Two scenario databases are frequently used in conjunction with the MESSEAGE model generator: IPCC RCP (Representative Concentration Pathways) scenario database and Greenhouse Gas Initiative (GGI) scenario database. Among them, the analyst chooses the appropriate baseline scenario to build their study on (IIASA III). For example, within the study on the effects of non-CO2 GHGs on climate change mitigation, Rao and Riahi (2006, p. 179) used the B2 scenario developed by the IPCC in the Special Report on Emissions. This scenario describes “local solutions to economic, social and environmental sustainability” (Rao and Riahi 2006, p. 179). MESSAGE facilitates modelling of all energy technologies. Moreover, the following characteristics can be accounted for: multiple inputs and outputs, seasonal variation &



64

storage, efficiency and costs varying with time, limits on production and resource extraction, nested and linked constraints, capacity build-up constraints, market penetration and environmental regulation (Rogner 2002, p. 25). A drawback in using MESSAGE, however, is the time consuming development of case studies, which has to be conducted with IAEA’s cooperation and can take several months.


MESSAGE is among57 the best suited models to address developing countries issues according to Urban et al. (2007, p. 3478). Characteristics such as electrification, traditional bio-fuels, urban-rural divide, structural economic change, subsidies, emission trading, clean development mechanism and renewable energies can be addressed (Urban et al. 2007; p. 3478, Table 2). MESSAGE focuses on the calculation of GHG related emissions. The main focus thereby lies on CO2 and CH4 and local pollutants like SOx and NOx. Further, Rao and Riahi (2006, p. 179) applied MESSAGE to calculate the impact of non-CO2 GHG, more specifically, of all six Kyoto GHG emission (CO2, CH4, N2O, HFCs, PFCs and SF6) on climate change mitigation. Cost-effective targets of GHG emission limits and reduction options are specified (Connolly et al. 2010, p. 1072). According to Mr. Jalal, Senior Energy/Nuclear Power Planner in the Planning and Economic Studies Section, Department of Nuclear Energy at the IAEA, MESSAGE is a very flexible tool that enables the construction of energy models with varying details. The wideness of the model in fulfilling the requirements necessary for the PROMITHEAS-4 mitigation/adaptation analysis will depend on how the model is constructed. The necessary cost accounting (calculation of mitigation/adaptation costs, social costs, etc.) can be conducted using MESSAGE, however, the analyst has to induce the model to do so. Also, the penetration rates of different technologies/sources can be outputs of the model.


The model package is provided free of charge to the public sector, non-profit organisations and research organisations, i.e. academia (COMMEND).


MESSAGE, was prominently used in the Environmentally Compatible Energy Strategies Project at IIASA, in cooperation with the World Energy Council, to assess the implications of global economic development on energy and environmental impacts (Messner and Strubegger 1995b, p. 10). As was mentioned under criterion a., it is used within the INPRO project by the IAEA member states as well as the European Commission.


IAEA provides in-depth training courses for IAEA member states taking approximately 2 weeks for basic applications (Connolly 2010, p. 1072).

57 Together with LEAP, RETScreen and WEM.



65

Conclusion

This report analyses the previously presented six models, ENPEP-BALANCE, MARKAL/TIMES, MERCI, LEAP, IMAGE and MESSAGE, according to the following catalogue of criteria:

a. The choice will be restricted to models used at European level;

b. The wideness of the model in covering mitigation/adaptation issues (The model that is closer

in covering these issues will be taken into consideration);

c. Transparency, complexity and easiness in using the model;

d. Availability of inputs (available in statistics books, national accounts);

e. Flexibility of the model in building scenarios (e.g.: a simulation model does not impose bias in

modelling outputs);

f. Compliance of outputs with our contractual obligations (socio-economic, technological

penetration);

g. Cost of acquiring the model;

h. International recognition of the model (used by governments);

i. Training and technical support.

Below, we summarize and list the models according to their capabilities to comply with the relevant criteria. The first question regards the use of the models at the European level. All models comply with this requirement. ENPEP(-BALANCE) was already used in many PROMITHEAS-4 partner countries. MARKAL/TIMES was applied in several EU projects and various European countries. MERCI is currently under construction and was used so far only in Austrian projects. The wide spectrum of countries applying the LEAP model within policy research projects includes European countries such as Greece, Estonia, Albania and Moldova. Within a number of projects conducted by the European Community, such as the ADAM (Adaptation and Mitigation) project, scenarios were developed with IMAGE. MESSAGE is used in a number of countries, mostly IAEA member and partner countries. Summarising, all the models seem to be widely accepted within the European community, noting that MERCI is still in the development process and has only been implemented on a national scale so far. An especially important subject from the PROMITHEAS-4 perspective is the wideness of the model in covering mitigation/adaptation issues. After having presented in detail the possibilities of IAMs to take into account mitigation, but more importantly the difficulties with modelling adaptation in a separate section, the best suited model to account for both is LEAP. However, all other models can include mitigation in their investigations. Moving on to the next topic, the transparency, complexity and easiness of the tools, an informal ranking can be constructed: MARKAL/TIMES is the most complex, however, transparent model for well trained analysts, followed by IMAGE, which requires a lot of expert knowledge. While MERCI is in the development phase, its usage is open only to analyst in collaboration with the developers. MESSAGE can well be categorised as an intermediate tool as far as transparency and complexity are concerned. ENPEP-BALANCE and LEAP, two simulation models, are the most transparent and easy to use tools in our set. Both have an intuitive user interface for fluent scenario construction.



66

The analysis of country specific mitigation and adaptation policies needs a solid data base. The models, however, vary quite significantly regarding the specifics necessary for a sensible investigation. MARKAL/TIMES, MESSAGE, MERCI and ENPEP-BALANCE have medium to extensive data requirements, while the data needs for LEAP and IMAGE will depend on the analysis the user is interested in. Moreover, there are databases available to be used especially in connection to LEAP and IMAGE respectively. It is evident that the set of models consists of powerful tools for scenario development. They will vary, however, in terms of flexibility in building these scenarios. To aid the decision process, also here an informal ranking is provided: Somewhat less flexible tools are MARKAL/TIMES and MESSAGE. ENPEP-BALANCE, MERCI and IMAGE offer more flexibility. The most flexible tool is LEAP, based on how easy scenarios can be adapted. Regarding the question, how these models take into account our contractual obligations, i.e. socio-economic aspects, social costs, administrative costs, costs for target groups, technological penetration rates and emerging/developing countries characteristics, the model that is best suited for this task is, again, LEAP. It is possible to account for a high number of developing countries specifics using LEAP, which will facilitate our analysis. Moreover, the model is highly flexible in addressing and calculating the relevant costs. MESSAGE also corresponds well to these criteria, followed by MARKAL/TIMES and IMAGE. MERCI and ENPEP-BALANCE can especially consider technological penetrations, but are useful tools for calculations of the other above mentioned aspects as well. Of course the respective cost of the models plays a crucial role, since the licence has to be acquired for twelve countries. MARKAL/TIMES, as the most sophisticated model in our set, is also the most costly one (within the range of €1.200 and €3.000). LEAP is free for developing countries; costs arise, however, for OECD countries. MESSAGE and ENPEP-BALANCE can be obtained free of charge. For MERCI no costs were specified so far. Developers of IMAGE, since it is only available in close cooperation with the developers, do not determine any costs for the use of this model. The most renowned model is definitely MARKAL/TIMES, while LEAP is fast approaching. However, all the other models, except for MERCI, have found wide international recognition. Turning to the last issue, the availability of training and support has to be evaluated. Although broad support and training are available for MARKAL/TIMES-users, the costs of both are high: basic training cost can amount up to €30.000. Training and support for the ENPEP-BALANCE model is available at €7.000 respectively. LEAP training is provided for free (expenses for the trainers have to be covered), support can be obtained against a fee. Costs for training and support for MESSAGE (training is conducted by the IAEA), IMAGE and MERCI are not specified. The striking aspects of this summary are, of course, the many advantages of LEAP for the purposes of the PROMITHES-4 project. The choice therefore, is not a difficult one and we can finalise this report with the conclusion that LEAP will be used for mitigation and adaptation analysis within the PROMITHEAS-4 project.

PROMITHEAS-4: “Knowledge transfer and research needs for preparing mitigation/adaptation policy portfolios”

67

Table 1. [Source: Author]

Use at

European

Level

Ability to

cover M/A

issues

Transparency,

Easiness

Required

Data

intensity

Flexibility

in

building

scenarios

Compliance

with

contractual

obligations

Cost International

recognition

Training,

Technical

support

(cost)

ENPEP-

BALANCE High Moderate High Moderate Moderate Moderate Low Moderate Moderate

MARKAL/

TIMES High Moderate Low High Low Moderate High High High

MERCI Low Moderate Moderate Moderate High Moderate Low Low Moderate

LEAP High High High Low High High Low High Low

IMAGE High Moderate Moderate Low Moderate Low N/A Moderate Not

specified

MESSAGE Moderate Moderate Moderate Moderate Low High Low Moderate N/A



68

Definition of Terms

Bottom-up models

This category encompasses models that “represent reality by aggregating characteristics of specific activities and processes, considering technological, engineering and cost details” (IPCC 2007, Annex I, p. 810). Top-down models

This are models “applying macroeconomic theory, econometric and optimization techniques to aggregate economic variables. Using historical data on consumption, prices, incomes, and factor costs, top-down models assess final demand for goods and services, and supply from main sectors, such as the energy sector, transportation, agriculture, and industry. Some top-down models incorporate technology data, narrowing the gap to bottom-up models” (IPCC 2007, Annex I, p. 821). Accounting-type models

Models applying an accounting framework consider flows of energy in a system determined through simple engineering relationships. The user determines all the relevant technology parameters and their values. The model is primarily a ‘sophisticated calculator’, whereby it can present various technology allocation scenarios (UNFCCC 2006, p. 15).



69

References

Balabanov T., Miess, M., Schmelzer, S. (2010). The Austrian Hybrid Dynamic Model E3: Methodology, Application and Validation, IHS WP-10-12 Balabanov T., Friedl B., Miess M., Schmelzer S. (2010). More and Better Green Jobs - Green Jobs for a sustainable, low-carbon Austrian economy. Final Report. Bellona Foundation (2011). A Bridge to a Greener Greece, [Can be downloaded at <http://cdn.globalccsinstitute.com/sites/default/files/fil_BELLONAprintFINAL.pdf>] Bhattacharyya S. C. and Timilsina G. R. (2010). A review of energy system models. International Journal of Energy Sector Management, Vol. 4, No. 4, 494 -518 Böhringer C., Rutherford T.,F. (2008). Combining bottom-up and top-down, Energy Economics volume 30, March 2008: 574-596 CEEESA- Center for Energy, Environmental, and Economic System Analysis, Argonne National Laboratory (2008). Energy and Power Evaluation Program (ENPEP-BALANCE), Brief Model Overview – Version 2.25.http://www.dis.anl.gov/pubs/61083.pdf CEEESA, Argonne National Laboratory. Overview of the Energy and Power Evaluation Program (ENPEP-BALANCE) (published without date) <http://www.dis.anl.gov/pubs/61124.pdf> Connolly D., Lund H., Mathiesen B.V., Leahy M. (2010). A review of computer tools for analysing the integration of renewable energy into various energy systems, Applied Energy 87 (2010) 1059-1082 Davis C. & Associates (2010a). Jamaica’s Greenhouse Gas Mitigation Assessment. Final Report. Davis C. & Associates (2010b). Greenhouse Gas Mitigation Assessment for Domenica. Final Report Dementjeva N. (2009). Energy Planning models Analysis and Their Adaptability for Estonian Energy Sector. Tallinn, TTÜ Press Heaps Ch. (2008). An introduction to LEAP. SEI – Stockholm Environment Institute. COMMEND – Community for Energy, Environment and Development. Heaps Ch. (2011). Quick Start Guide for Using Optimization in LEAP. SEI. COMMEN. [Source: http://www.energycommunity.org/documents/OptimizationQuickStart.pdf] Heaps Ch., Erickson P., Kartha S., Kemp-Benedict E., SEI (2009). Europe’s Share of the Climate Challenge. Domestic Actions and International Obligations to Protect the Planet. Hoogwijk M., Rue du Can S. de la, Novikova A., Urge-Vorsatz D., Blomen E. (2008). Assessment of bottom-up sectoral and regional mitigation potentials. Background report. MNP. Scientific Assessment and Policy Analysis programme for climate change (WAB). IEA – International Energy Agency/ETSAP – Energy Technology Systems Analysis Programme (2008). Global Energy Systems and Common Analyses, Final Report of Annex X (2005-2008), [Gary Goldstein and GianCarlo Tosato (eds)]



70

IPCC-Intergovernmental Panel on Climate Change (1996). Technologies, Policies and Measures for Mitigating Climate Change. IPCC Technical Paper I. IPCC (2001). Climate Change 2001: Mitigation. Contribution of Working Group III to the Third Assessment Report of The Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. IPCC (2007). Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, [B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. King B.B. (1985). What is a SAM?. Social Accounting Matrices. A Basis for Planning. edited by Graham Pyatt and Jeffery I. Round. p. 17 – 51. Washington D.C.: The World Bank

Loulou R., Remne U., Kanudia A., Lehtila A., Goldstein G. (2005). Documentation for the TIMES Model PART I. ETSAP, http://www.etsap.org/documentation.asp Messner S., Strubegger M., (1995a). User’s Guide for MESSAGE III. WP-95-69. International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria. Messner S., Strubegger M., (1995b). Model-Based Decision Support in Energy Planning. WP-95-119. International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria. Miess M., Schmelzer S., Schnabl A. (2011). Evaluation of the regional employment and growth offensive 2005/2006. A quantitative and qualitative Analysis. Forthcoming. MNP (2006). (Edited by A.F. Bouwman, T. Kram and K. Klein Goldewijk), Integrated modelling of global environmental change. An overview of IMAGE 2.4. Netherlands Environmental Assessment Agency (MNP), Bilthoven, The Netherlands. Pyatt G. (1988). A SAM approach to modeling. Journal of Policy Modeling Vol. 10, p. 327 - 352

Rao S. and Riahi K. (2006). The Role of Non-CO2 Greenhouse Gases in Climate Change Mitigation: Long-term Scenarios for the 21st Century. The Energy Journal Reinert K. A., Roland-Holst D. W. (1997). Social Accounting Matrices. Applied Methods for Trade Policy. A Handbook. edited by Joseph F. Francois and Kenneth A. Reinert, p. 95 - 120, Cambridge: Cambridge University Press.

Rogner H. H. (2002). IAEA Tools for Integrated Energy System Analysis. Planning and Economic Studies Section, International Atomic Energy Agency (IAEA) [Source: <http://www.energia.inf.cu/eventos/memorias2/evento/cuba_july02_1.pdf>] van Ruijven B., van Vuuren D.P: (2009). Oil and natural gas prices and greenhouse gas emission mitigation. Energy Policy. Vol 37(11). 4797-4808



71

SEI - Stockholm Environment Institute (2011). User’s Guide for Version 2011. First Draft. [Source: http://www.energycommunity.org/documents/LEAP2011UserGuideEnglish.pdf ] SEI (2006). Data Requirements for Energy Planning and Mitigation Assessment. [Source: http://www.energycommunity.org/documents/DataRequirements.pdf ] UNFCCC-United Nations Framework Convention on Climate Change (2006). Module 5.1 – Mitigation Methods and Tools in the Energy Sector <http://unfccc.int/resource/cd_roms/na1/mitigation/index.htm> [accessed: 11/05/2011] Urban F., Benders R.M.J., Moll H.C. (2007). Modelling energy systems for developing countries. Energy Policy 35 (2007), 3473-3482.



72

URLs:

CEEESA (I) - Argonne National Laboratory. List of ENPEP applications in Europe <http://manhaz.cyf.gov.pl/manhaz/links/EMISSION%20TRADING/Applications%20of%20the%20Energy%20and%20Power%20Evaluation%20Program%20%28ENPEP%29%20in%20Europe.htm> [accessed 26/05/2011] CEEESA (II) – world wide applications list <http://www.dis.anl.gov/news/EnpepwinApps.html> COMMEND (I) – Community for Energy, Environment and Development http://www.energycommunity.org/default.asp?action=71 COMMEND (II). http://www.energycommunity.org/default.asp?action=72 GAMS-L - GAMS users worldwide mailing list http://www.gams.com/maillist/gams_l.htm GAMS-SALES - http://www.gams.com/sales IAEA – http://www.iaea.org/INPRO/index.html IEA – International Energy Agency/ETSAP – Energy Technology Systems Analysis Programme (I), Guidelines http://www.etsap.org/TOOLS/ETSAP_SW_Guidelines.pdf IEA/ETSAP (II) – Tools description http://www.etsap.org/Tools.asp IIASA (I) – International Institute for Applied Systems Analysis (I) http://www.iiasa.ac.at/Research/ENE/model/extensions.html IIASA (II) - < http://www.iiasa.ac.at/Research/ENE/model/message.html> IIASA (III) - http://www.iiasa.ac.at/Research/ENE/GGIDB_index.html and http://www.iiasa.ac.at/web-apps/tnt/RcpDb/dsd?Action=htmlpage&page=welcome MNP (I) – Netherlands Environmental Assessment Agency http://themasites.pbl.nl/en/themasites/image/projects/reports/ensembles.html MNP (II) – http://themasites.pbl.nl/en/themasites/image/projects/articles/index.html MNP (III) - http://themasites.pbl.nl/en/themasites/image/projects/reports/index.html MNP (IV) - http://themasites.pbl.nl/en/themasites/image/overview/index.html NEEDS - New Energy Externalities Developments for Sustainability (NEEDS) Project http://www.needs-project.org/index.php?option=com_frontpage&Itemid=1 RES2020 – Renewable Energy Sources – Project <http://www.res2020.eu/files/fs_inferior01_h_files/pdf/deliver/The_PET_model_For_RES2020-110209.pdf>

Communication

The following model developers were contacted for further information: Gary Goldstein – MARKAL/TIMES Charles Heaps - LEAP Ahmed Irej Jalal - MESSAGE Detlef van Vuuren – IMAGE

Deliverable D - UoA...PROMITHEAS-4: “Knowledge transfer and research needs for preparing...

Documents

Transcript of Deliverable D - UoA...PROMITHEAS-4: “Knowledge transfer and research needs for preparing...