Model-based analysis of the future strategies for the MENA energy system

Transcription

1 Energy Strategy Reviews xxx (2012) 1e12 Contents lists available at SciVerse ScienceDirect Energy Strategy Reviews journal homepage: ANALYSIS Model-based analysis of the future strategies for the MENA energy system Panagiotis Fragkos, Nikos Kouvaritakis, Pantelis Capros * National Technical University of Athens, Department of Electrical and Computer Engineering, 9 Iroon Politechniou Street, Zografou Campus, Athens, Greece A R T I C L E I N F O A B S T R A C T Article history: Received 28 September 2012 Received in revised form 7 December 2012 Accepted 14 December 2012 Available online xxx Keywords: Energy strategies Energy modelling Energy policy This paper introduces a large-scale energy demand and supply model that is used to quantify alternative energy system strategies for the Middle East and North Africa (MENA) region to MENA contains major hydrocarbon producers and a vast and currently untapped potential for renewable power generation. It examines mutual benefits that MENA and the EU could derive by cooperating in the field of energy and climate policies. It also investigates a strategy emphasising decentralised RES deployment together with accelerated market reform leading to a reduction in power generation costs and a large increase of exportable hydrocarbon surpluses. Recognising the risks that characterise the region a case of policy failure is also considered. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction 1 The Middle East and North Africa (MENA) countries face major energy policy challenges in rapidly changing situations. Fossil fuel producing countries are concerned with an optimal intertemporal management of their ultimately finite resources, while net importers of fossil fuels aim at reducing their dependence in order to decrease their vulnerability and improve their economic development prospects. The region contains countries that fall in both categories. At the same time most of them are characterised by a very large potential for expansion of the use of renewable energy sources. CO 2 emissions in MENA have been growing very fast and there are strong indications that costeffective abatement options exist opening the way for collaboration with the EU where climate policy is high on the agenda. In view of these policy goals, there has been a growing interest with respect to the development of the Mediterranean energy system, aiming at determining the optimal future mix of electricity generation and specifically addressing the central question of deployment of renewable or nuclear energy. Marktanner and Salman [4] focus on the wider economic and geopolitical impacts of large-scale development of * Corresponding author. address: kapros@central.ntua.gr (P. Capros). 1 We acknowledge research funding by the European Commission within the framework research program MEDPRO ( RES and nuclear technologies in North Africa, while Brand and Zingerle [5] summarise the renewable energy targets of the Maghreb region and then employ a linear power market optimisation model to assess the impact of the accomplishment of these targets on electricity supply costs. Supersberger and Führer [6] present the effects of integration of RES and nuclear energy into North African energy systems for the region s balance of trade and highlight the fact that while RES deployment will allow North Africa to ensure independence from energy imports and to guarantee fossil exports for a longer period, in the case of nuclear development the North African countries will strongly depend on fuel and technology imports. Reflecting the increasing interest on the potential cooperation of Mediterranean countries on the fields of energy and climate action, several studies have been published recently (e.g. Trieb and Müller- Steinhagen [7], Folkmanis [8]). The MENA region has huge potential for solar and wind energy, which has remained to a large extent untapped. Boudghene-Stambouli [10] emphasises the vast RES potential of Algeria (mainly concerning CSP). Viebahn et al. [9] identify the CSP technology as the most cost-effective carbon-free power generation option for the MENA countries. In the last decade the concept of large-scale CSP deployment in North African countries together with the potential export of green electricity to the EU through high-voltage directcurrent (HVDC) lines has attracted a growing attention. Both national regulatory authorities and international (mainly European) private or governmental initiatives, like the Desertec Foundation, MEDGRID and EUROMED, have conducted feasibility studies and costebenefit analysis X/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.

2 2 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 [11,12,14] in order to assess the sustainability and the economic viability of this concept and project the future evolution of the EU- MENA power supply system. Another strand in the literature investigates econometrically the long-term relationship between GDP per capita, energy consumption and CO 2 emissions. In Ref. [1] El Hedi Arouri et al. use a panel model to estimate the long run elasticities of CO 2 emissions with respect to energy consumption and real GDP and to explore the nature of the causality relationship between economic growth, energy consumption households, services and transportation. These activity forecasts, together with consumer prices for fossil fuels which are themselves derived from international fuel prices (coming from the latest PROMETHEUS projections [19]), taking into account country specific characteristics, such as distribution costs, taxation and fuel subsidies, are used to determine the evolution of energy demand by sector. Long-term and short-term price effects are accounted for separately by using different elasticities. The general equation below describes the main mechanism for determining the evolution of final energy demand ðc i;t Þ in each sector i at time t a ACTVi;t APi;t b1* APi;t 1 b2* Y n C i;t ¼ C i;t 1 * * ACTV I;T 1 AP i;t 1 AP i;t 2 k ¼ 1 APi;t k AP i;t k 1 4ðk=nÞ*g (1) and emissions of CO 2. In Ref. [2] Al-Mulali examines the impacts of oil consumption on economic growth in the MENA region. The literature is much poorer with regard to the development of model-based energy system analysis for the MENA countries apart from the power generation sector. Only IEA [13] and the Mediterranean Energy Observatory (OME) [15] quantify energy demand and supply projections for the entire energy system. OME uses an econometric model to simulate the evolution of final energy demand, power generation mix and primary fuel supply for the Mediterranean countries. This paper presents a detailed 2 and technology rich integrated demand and supply model for the countries of the region (Section 2). This model has been used to analyse contrasting and policy induced demandesupply configurations and the main results of the analysis are presented in the remainder sections. In view of the rapidly evolving social and political context in the region, the emphasis of the policy cases examined is placed on alternative positions with regard to integration in the world economy and the development of bilateral and multilateral cooperation and their implications on the entire energy system. 2. Methodology e model description MENA-EDS 3 is a large-scale energy demand and supply model 4 that simulates the formation of prices in energy markets, estimates the quantities demanded and supplied by the main energy system actors in an exhaustive manner and incorporates energy related CO 2 emissions, environmentally oriented policy instruments and emission abatement technologies. MENA-EDS is designed for medium-term and long-term projections and produces analytical quantitative results for each country until Historical energy demand and supply data for the years up to 2010 are derived from the IEA and ENERDATA databases. The model has been applied to the MED-9 region that includes Algeria, Morocco, Tunisia, Egypt, Libya, Israel, Syria, Lebanon and Jordan. It calculates CO 2 emissions and energy system costs (including power generation costs) and can simulate energy strategies and policy instruments such as taxes and subsidies, carbon pricing mechanisms and incentives promoting energy efficiency and renewable sources. MENA-EDS is a recursive dynamic model with annual resolution and has a predominantly triangular structure in order to limit contemporaneous simultaneity. On the other hand, simultaneity is modelled through lagged instances of endogenous variables. The MENA-EDS model takes as an exogenous input demographic, macroeconomic and sectoral activity projections, covering the major energy consuming sectors in industry, 2 The model contains approximately 6000 equations for each country. 3 Middle East and North Africa Energy Demand and Supply model. 4 E3MLab constructed and operates the model. where 4ðk=nÞ are the weights of the polynomial distributed lag used to estimate the long-term price elasticity in each subsector, ACTV i;t is the activity indicator for each sector (either GDP, sectoral value added, disposable income, tn-km or passenger-km), AP i;t is the average energy price for sector i a represents the activity elasticity, b 1 and b 2 the short-term price elasticities that capture demand reactions to price that do not require significant investments and g the long-term price elasticity reflecting energy saving investments over a longer period. All the above parameters have been estimated econometrically and 4ðk=nÞ is a second order polynomial distributed lag scheme of depth n with near and far zero restrictions. Final energy demand is simulated for five main sectors: Industry, where ten sub-sectors are included in the analysis depending on data availability Services Households Agriculture Transport (including private passenger cars, road freight transport, passenger aviation and rail transport depending on data availability) Energy demand arises from net increases in consumers energy needs, the replacement of scrapped capacity and surviving equipment. Regarding new energy demand, a compact but analytically rich specification encapsulates the dynamic process of technological substitution in all sectors taking into account the technical and economic characteristics (investment costs, energy efficiency, fixed and variable operating costs) of the technologies available. The evolution of the passenger car stock in different countries is simulated in detail by considering the effect of economic development and behavioural changes on both the number and the use of vehicles, also allowing for potential saturation effects. A detailed representation of the power supply sector has been implemented in the MENA-EDS model, as electricity generation is projected to play an increasingly important role in energy and climate policies. Generation requirements are determined by final electricity demand, own-consumption of power plants, electricity trade between countries and transmission and distribution losses in each country. The sectoral origin of electricity demand is used to construct an annual load duration curve, by taking into account that demand in energy intensive industrial sectors is mainly base load, while pronounced peaks characterise demand in services and households. A wide variety of technological options compete to satisfy electricity demand. The main categories of power generation options are: Gas-fired technologies, using steam turbine, gas turbine or combined cycle technology

3 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 3 Coal-fired technologies, which include thermal, fluidised bed, supercritical and integrated gasification technologies using coal or lignite as a fuel Oil-fired technologies, including thermal fuel oil and peak devices fuelled by diesel Nuclear technologies, third and fourth generation Biomass-fired technologies, including thermal and integrated gasification technologies A wide variety of renewable technologies, including hydroelectricity (large or small scale), wind (onshore and offshore), solar (CSP, 5 photovoltaic) and geothermal Capacity installation decisions are based on long-term marginal costs (that include the annualised and discounted capital costs, the fixed and variable operation and maintenance costs and fuel costs) in combination with expectations for the load duration curve. The market share of each technology in new investments for the year t is modelled as a quasi-cost minimisation function and is determined by the longterm marginal costs ðc j;t Þ of the competing options. share j;t ¼ w j;t*c g j;t Pj w j;t*c g j;t In this specification, the parameters w j;t can be interpreted as reflecting the relative economic and technical maturity of each technology while the parameters g represent the sensitivity of the market share with respect to the total cost of each technology. For the calculation of long-term marginal costs the technical and economic characteristics of the PRIMES model [18,20] are used, taking into account regional specificities, such as the higher capacity factors for CSP and photovoltaics. Scrapping rates of power plants are endogenous in the model and include both normal scrapping, due to plants reaching the end of their lifetime, and premature scrapping, due to changes in variable and fuel costs which render the continuation of a plant s operation economically unsustainable. The model also accounts for already decided investments in specific power plants and the plans adopted for decommissioning of old and inefficient ones, as obtained from a wide variety of literature review. The annual load duration curve together with variable operating costs and the installed capacity of the different technologies are used to determine capacity utilisation for each time segment (dispatching of power plants) and hence electricity production and associated fuel inputs for each technology. Firstly, the year is divided into 9 h segments, which are symbolised by the index u, u ¼ 0,.,8. The annual load duration curve is approximated by a rectangular section representing base load and an exponential section accounting for the shorter durations. Total electricity production for the year t (TOTPROD t )is then approximated by the following formula: TOTPROD t ¼ X8 u ¼ 0 (2) h ðm t B t Þ*e l t*ð0:25þuþ i þ 9*B t (3) where M t is the highest load demand considered, B t is the base load demand and the parameter l t is calculated implicitly from the equation: 1 e 8:76*l t l t ¼ PROD t 8:76*B t M t B t (4) where PROD t represents electricity generation. The price of electricity is determined as a function of long-term average marginal costs and is differentiated between the sectors (industry, residential), reflecting the differential costs for each sector (differences mostly arise from the fact that different technologies supply different segments of the load duration curve and from differential distribution costs). Any taxes and subsidies are added exogenously and constitute policy instruments. Primary production of fossil fuels is a function of reserves, investments in productive capacity and, in the case of gas and coal, demand (both internal and for export). For crude oil it is assumed that the world market can absorb whatever quantities can be produced. Reserves are determined by a motion equation that calculates net additions in terms of discoveries minus production. The rate of discovery depends on fuel prices and the undiscovered resources of the fuel as estimated by geologist experts [3]. The difference between primary consumption and primary production gives net trade through an identity. Natural gas trade between countries takes into account existing pipeline and LNG infrastructure and projects their future evolution. 3. The policy alternatives The MED-9 region has been over many decades in a state of political, social and economic flux. There are as yet no clear indications that this situation is changing. Consequently, any projection concerning the region is subject to greater uncertainty than is usually the case with medium to long-term projections. Clearly the evolution of its energy system will depend largely on these uncertainties. The MENA-EDS model has been used in order to address questions arising from possible directions of energy policy in the various countries covered by the model. Such policies will to a large extent depend on the international context within which the region may evolve and in particular the degree of cooperation between the countries themselves and their integration in the wider regional and global economic systems. The analysis involves the examination of alternative courses that such cooperation and integration may follow in the two coming decades. The results of the present work are presented in terms of comparison of alternative quantified projections primarily with a reference projection. The latter assumes a very gradual normalisation of the region with energy policies and measures that are currently on the political agenda of the different countries implemented at varying but generally cautious rates. Since this projection is used as the benchmark for comparisons it is presented separately and briefly on Section 4. This presentation aims to introduce exogenous assumptions on the most important drivers and trends for key variables that characterise the energy system of the region. The Mediterranean region is of strategic importance to the EU both in economic and political terms. In the 1995 meeting in Barcelona, the EU committed itself to promoting Euro-Mediterranean economic and political cooperation and at the July 2008 summit decided to upgrade the Barcelona Process and to create the Union for the Mediterranean. 6 The Mediterranean and the EU countries cooperate in the field of regional environmental protection and sustainable development [16]. These initiatives indicate that there is political will for cooperation across the Mediterranean. It is clear however that there is plenty of scope for deepening relationships and in particular engaging MED-9 countries in EU climate policy efforts. The MED-EU Initiatives strategy assumes that projects such as Desertec will to a large extent materialise and that the EU Emissions Trading Scheme (ETS) will expand to include MED-9 with special provisions for these countries (free allocation of permits to the extent of reference projection emissions). Only energy intensive industries and power generation are subject to 5 Concentrated Solar Power. 6 See: /sommet_de_paris pour_la_ mediterranee_4758.html.

4 4 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 Table 1 Main scenarios assumptions. Reference MED-EU Initiatives Global Integration Fragmentation Energy price reform Gradual, incomplete by 2030 Accelerated, especially in industry and power generation, incomplete by 2030 Electricity exports No 235 TWh in 2030 to the EU, assisted by EU feed-in tariffs RES policies Only policies that are already firmly decided Efficiency standards Present policies Present policies gradually strengthened Accelerated, complete by 2025 Limited depending on costs Continuation of present energy subsidies RES facilitating policies RES facilitating policies Delays in present plans Additional standards, vigorously Present policies but gradually introduced ETS carbon prices No For industry and power generation No No Cooperation Continuation of the With the EU primarily in the fields Multilateral No present situation of climate policy and RES promotion Investment climate Gradual reduction of risks Lower risk premiums leading to higher capital turnover and more FDI, mainly industry and power generation Much improved with lower perceived risk vis a vis the rest of the world and internally Continuation of the present situation No allowance obligations. The MED-9 countries do not face an additional cost associated to emissions, but may benefit from an opportunity of generating revenues by reducing their emissions and by selling their allowances to EU countries. Such sales are projected to amount to 64.6 bn V 08 cumulatively between 2012 and Apart from direct sales of emission permits, the exploitation of emission reduction opportunities may result in benefits for MED-9 countries arising from an increase in foreign direct investment (FDI). This prospect implies that renewable facilitating policies (licensing and others) accompany the ETS enlargement. Renewable electricity exports to the EU will require investments into new electricity high-voltage DC interconnectors linking the MED-9 countries with the southern countries of Europe. The exported renewable electricity is charged at pre-defined fixed tariffs which are set at a sufficient level to allow recovery of total capital and operating costs and allows for a reasonable rate of return on capital (at an 8% discount rate). The cumulative value of electricity exports between 2020 and 2030 amounts to 85.9 bn V 08. The MED-EU Initiatives strategy assumes cooperation and actions affecting essentially emissions from energy intensive activities and in particular the power generating sector with few spillovers to other sectors. On the other hand, there is a huge potential for improvements on a very wide front addressing challenges such as economic and energy reforms, trade liberalisation, infrastructure upgrading and sustainable development. The Global Integration strategy assumes that MED-9 countries individually undertake vigorous measures in order to promote energy efficiency, the development of renewable energy sources, a reduction of import dependence for net importers of energy and enhancement of the export capability of the energy exporting countries. More specifically price reform is accelerated, grid improvements are brought forward and additional measures promoting efficiency standards are introduced. It also assumes that relations of individual MED-9 countries with the rest of the world deepen and as a consequence perceived risks are diminished thus encouraging FDI. On the other hand, the promotion of renewable electricity exports to Europe is assumed to be at a much more limited scale compared to the massive effort assumed in the MED-EU Initiatives strategy. The Global Integration strategy is characterised by multiple decentralised actions, often at a small local scale, that taken together constitute a major attempt at transforming the energy system. Naturally, for such conditions to prevail within the forecast horizon, very rapid political normalisation must occur and there are currently no clear signs that this is happening. It is considered here primarily in order to chart the potential for rationalisation of the energy system of the region. The MED-9 area in the Fragmentation case is characterised by increased fragmentation, sporadic and festering conflicts and a failure of concerted action with the EU and other global players. In terms of the energy economy the main implications are a shortage of capital, increased investment risks leading to high risk premiums and a stalling of market reform including the price reform that is assumed to be introduced gradually in the reference. Under these circumstances of course there is no scope for investing in interconnections to enable the export of renewable electricity to the EU. The perceived risks assumed in this Fragmentation case are not evenly associated with the different countries and they depend on their current condition and their recent history thus they are most strongly felt in countries like Syria and Lebanon and less acute in countries like Israel and Tunisia. Table 1 summarises the main assumptions of the four scenarios examined. 4. The reference projection 4.1. Present situation e summary MED-9 consists mostly of emerging economies that are characterised by different stages of development. These differences are to a certain extent reflected in the primary energy demand and electricity production per capita indicators. In most cases there is clearly a large scope for MED-9 countries to increase their consumption per capita in line with economic development and standards of living and comfort. On the other hand, the region is characterised by a different climate than EU-27 and need not necessarily converge at saturation levels comparable to Europe. Looking at dynamic trends as they are reflected by crude elasticity measures of energy demand with respect to GDP it is worth noting that over the period 1990e2010 the region as a whole has registered a value that is very close to unity compared to a mere 0.12 for EU-27. Over the same period the empirical elasticity of power generation with respect to GDP for the region has been more than double the equivalent for EU- 27 (1.59 instead of ). In general electricity demand in the MED-9 countries shows little sign of reaching saturation levels in the medium term Assumptions for the reference projection The MED-9 region as a whole has been characterised by relatively high rates of population growth in the recent past. The reference 7 Calculated by the authors using historical data from the IEA and ENERDATA databases for the MED-9 region and EUROSTAT for the EU.

5 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 5 Table 2 International fossil fuel prices in $2010/boe. a Oil price Gas price Coal price a Barrel of oil equivalent. Table 3 Shares (in %) of energy forms in final energy demand of industry in the MED-9 region Solids Oil Gas Electricity projections are derived from the medium fertility variant of the World population prospects of the United Nations [17]. This projection implies a marked deceleration in population growth for all countries in the region. This trend is particularly pronounced in Jordan, Lebanon, Israel and Syria. The reference projection also assumes a continuation of the trends in terms of urbanisation with marked consequences for changes in lifestyles and energy consumption patterns. Economic activity growth in MED-9 countries over many years has been strongly influenced by political instability. Nonetheless in the period from 1990 to 2010 growth has on average occurred at high rates. For the reference projection a certain amount of political normalisation is assumed leading to increased trade and cooperation. Another driver for energy demand is the evolution of consumer prices for energy. In principle movements in international fuel prices 8 (Table 2) should be reflected in domestic consumer prices. However, the MED-9 region is characterised by a great variety of pricing regimes. Looking at transportation fuels, Israel and Morocco have prices and taxation comparable to the prices that prevail in the EU. In Tunisia and Lebanon transport fuel taxation is very low. The other countries in the region effectively subsidise transport fuels since prices for the consumer are lower than tax-free spot prices for exports. Clearly this situation makes little economic sense since such spot prices (i.e. the prices at which the fuels could be sold if they are not consumed locally) constitute an opportunity cost. The reference projection assumes a very gradual movement towards rational transport fuel prices in these countries. Concerning electricity prices, it is generally agreed that in order to have a sustainable generation and distribution system consumer prices must cover long-term marginal costs (i.e. apart from operating costs they should also include appropriate capital annuities to ensure that investments will be profitable). Industrial consumer prices in Jordan, Syria and Egypt do not cover long-term marginal costs calculated using an 8% discount rate. In the case of Algeria and Libya industrial prices are particularly low but they cover long-term marginal costs because of the extremely low prices of natural gas inputs. Residential prices in the region cover long-term marginal costs only in Tunisia and Israel. Even in countries like Morocco where energy price reform has progressed significantly residential electricity prices are still subsidised albeit to a lesser extent. The usual justification for such subsidies is that electricity uses in households effectively perform a social service in facilitating the enjoyment of material civilisation for every citizen. In the reference projection price reform is assumed to take place gradually and at a different pace depending on the country (more slowly for Algeria, Libya, Egypt and Syria) Industrial energy demand In the projection period due to the gradual opening up of the economies and a shift in domestic demand towards services, the share of industrial value added in GDP is forecast to follow a declining trend (between 2010 and 2030 the average industrial growth is 2.7% compared to almost 4% for GDP). On the other hand, the persistence of low energy prices in most MED-9 countries means that international specialisation will favour more energy intensive sub-sectors and as a consequence specific energy consumption of industry declines rather slowly. The projection implies a big increase in the share of electricity in total final energy demand for industry 9 as a result of a penetration of electrical industrial processes and increased demand for specific electricity needs, such as electric motors. Another striking feature of the projection is the increased penetration of natural gas for heat and steam raising purposes mostly at the expense of oil (Table 3). This substitution occurs primarily for purely economic reasons: with the expansion of the natural gas grid there is greater potential for more attractively priced gas to substitute residual fuel oil Energy demand in residential/commercial sectors The residential/commercial sectors currently account for around 30% of the MED-9 region s total final energy demand. Energy demand for space heating is not particularly important, while for cooking and water heating purposes there is a big variety of fuels used, including LPG, natural gas, traditional biomass and electricity. The use of traditional biomass is projected to decline continuously throughout the period because of increased urbanisation and rising living standards. LPG use is projected to lose share to the extent that households are connected to the natural gas grid. Natural gas is forecast to increase its share in total residential/commercial demand from 16% at present to 26% by 2030 (Table 4). Electricity demand, which is projected to increase 2.8 times from its 2010 level, is pushed forward by the rapid penetration of electric appliances, such as refrigerators and deep freezers, washing machines, television sets and many other appliances that in many countries of the region are far from having reached saturation levels. A special mention in this context concerns air-conditioners mostly used for cooling purposes. They already tend to modify the load curve in many countries Energy demand in the transport sector The major uncertainty with regard to energy consumption for transport in the region arises from the possible evolution of car ownership in the different countries (Table 5). At present most countries have very low car ownership rates [22]. Lebanon is an exception with a rate approaching 400 vehicles per thousand inhabitants. The model-based projections resemble to S-shaped penetration curves simulating take-off and saturation effects. There is an overall tendency for reduction in vehicle utilisation rates as motorisation increases. On the other hand, the share of urbandriving in conditions of congestion increases significantly over the forecast horizon in all countries of the region. The average size of vehicles will have a slight tendency to increase as incomes rise. Better 8 International fuel prices are derived from the latest PROMETHEUS model reference projection, carried out in Electricity demand for desalination purposes primarily through reserve osmosis is included in industrial demand. The projections have been introduced exogenously following [23].

6 6 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 Table 4 Shares (in %) of energy forms in final energy demand of residential/commercial sectors in the MED-9 region Solids Oil Gas Traditional biomass Electricity vehicle efficiency will occur through car imports. The modest pump price reform assumed for some of the countries will have a relatively small effect on vehicle choices. The net result of all the above is that fuel consumption per vehicle displays a modest improvement of 1.2% p.a. Commercial road transport in 2010 accounted for approximately 35% of total energy consumption in road transport. Energy consumption for trucks is closely linked to economic activity and is projected to increase vigorously (2.7% p.a.). Air transport activity is expected to show particular dynamism as the number of passengers carried is projected to grow by 5.1% p.a. over the forecast period, compared to a growth of 3.9% p.a. for GDP. Aircraft occupancy rates will slightly decline but improvements in the energy efficiency of aircrafts will mean that fuel consumption by the air transport sector will grow at an average rate of 4% p.a Electricity sector The sustained electrification of industry and the rapid penetration of electrical appliances in the residential sector translate into vigorous growth in final demand for electricity (Table 6) for most countries in the region (the main exception is Israel where saturation effects become apparent). Growth of demand for electricity is projected on average faster than the growth of GDP and means that by 2030 the region becomes a major electricity market (1042 TWh) requiring large-scale expansion of productive capacity (from 92 GW in 2010 to 243 GW in 2030). In all countries of the region (with the exception of Israel) there is considerable scope for transmission and distribution losses reduction. The reference scenario assumes their gradual but sustained reduction. 10 The nuclear option has at various times been considered by a number of countries including Algeria, Morocco, Israel and Jordan. The Fukushima accident has brought into the fore a considerable amount of scepticism concerning most of these projects. Even if some of them finally go ahead, experience from the past shows that given present concerns the whole process of planning, tendering, licensing and construction may take even more than the two decades that separate us from the projection horizon. The hydro-electric potential of the region has already to a large extent been tapped. A minor expansion of hydropower is projected for Morocco (it concerns 560 additional MW). Wind power is already being exploited in the region notably in Egypt, Morocco and Tunisia. Most countries have ambitious plans for increasing the contribution of wind power. The main instruments used for promotion are direct investment by state owned enterprises, investment subsidies, feed-in tariffs and quotas accompanied by 10 In the cases of Syria and Algeria recorded losses are so high that they clearly point to unrecorded and unpaid deliveries. The projection incorporates adjustments on demand in the household and services sectors in these countries in order to avoid biased electricity requirements. In this sense for these countries the figures for these sectors represent electricity purchases rather than consumption. Table 5 Evolution of private cars per thousand inhabitants in the MED-9 countries. ALG MOR TUN EGY LIB ISR LEB SYR JOR MED economic instruments. The reduction in capital costs of recent years together with the availability of many suitable sites (enabling high utilisation rates) in combination with the various promotion policies are likely to produce a large expansion of wind capacity in the coming two decades (Table 7). However, even so this represents a small fraction of the potential. With regard to solar thermal power production the main technology considered in the model is CSP with varying storage capacities. The region as a whole but especially its Saharan parts is considered to contain among the most suitable sites for this type of technology in the world. CSP is characterised by relatively high cost especially when compared with its most obvious competitor, which is the combined cycle gas turbine technology. Under these conditions CSP deployment necessitates a special support system that goes even beyond the one created in order to encourage wind power. In very recent years such support systems have started being put in place and there is considerable interest in the promotion of many CSP projects in a number of countries, such as Algeria, Morocco, Egypt, Israel and Jordan. The reference projection presented here does not incorporate large-scale exports of renewable electricity from MED-9 countries to the EU. However, it assumes that the present effervescence concerning the possibility of such exports leads to the undertaking of a number of projects and the creation of a suitable framework in which CSP is initially deployed. Photovoltaic generation is less systematically pursued than CSP. It involves mainly relatively small units and also requires considerable support to become competitive under the conditions prevailing in the energy markets of the MED-9 countries; another factor limiting wide development of PV is the lack of adequately meshed low/medium voltage grids and the relatively high investment that would be needed. Only two countries (Morocco and Israel) currently use coal for power generation and no other country is projected to use such options. The share of coal in power generation will substantially decrease in the forecast period, as no new investments in coal-fired power plants are projected for Israel since domestically produced gas fuels the majority of additional capacity. Oil as a fuel for power generation has seen its shares drop sharply in recent years. This process is forecast to continue in all countries until 2030 by which time many of them will generate virtually no oil-fired electricity. This rapid transition takes place in the light of competition from gas, facilitated by higher production and increased intraregional trade for the latter. Natural gas already dominates the power generating sector of the region and its position overall is projected to strengthen in the coming two decades. The dominant option for new gas-fired capacity is the combined cycle gas turbine technology. It combines relatively low capital costs with very high efficiency rates thus making it attractive even in natural gas importing countries where the prices of the fuel Table 6 Average annual growth of total final electricity demand (in %). ALG MOR TUN EGY LIB ISR LEB SYR JOR MED e

7 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 7 Table 7 Evolution of electricity production in the MED-9 region in the reference projection (in TWh) Electricity supply Nuclear Gas Oil Solids Hydro Wind Biomass & waste Solar approach international levels. Gas-fired power generation is also convenient in complementing intermittent renewable power production and facilitates load management especially when renewable shares are relatively high. 5. Analytical comparisons 5.1. Impacts on final energy demand In the MENA-EDS model, consumer prices are major drivers of final energy demand. The different strategies examined involve price reform aimed at economic rationality and providing appropriate signals to consumers that promote efficiency gains and the desired substitution between fuels. The scope and pace of such reform differs according to the strategy pursued and varies between countries depending on existing pricing practices. Fig. 1 below illustrates the extent of price reform as reflected on two key consumer prices (average prices for gasoline at the pump and electricity to residential consumers). The averages shown exclude Israel, Tunisia and Morocco where the need for reform is less urgent. It is worth noting that with the exception of the Fragmentation projection, the strategies imply a clear and early break with the recent past. Nevertheless prices in the region remain generally lower than those currently prevailing in the EU even for the proactive strategies. In the MED-EU Initiatives strategy energy intensive industry is specifically targeted (participates in the common EU-MED-9 ETS). As a consequence final demand in industry is strongly affected. This is particularly pronounced in countries like Algeria and Libya, where fuel prices in the reference projection are very low, implying a very large increase in fuel costs. On the other hand, countries like Tunisia, Israel and Morocco, where fuel prices are high even in the reference projection, the impact of the MED-EU Initiatives strategy is relatively small. In Algeria, Tunisia, Egypt and Libya, where natural gas in the reference projection overwhelmingly dominates industrial energy demand, substitution possibilities are more limited. In the Global Integration strategy the industrial sector has a more limited potential for efficiency gains compared to residential uses. This is mainly due to the fact that industrial consumers who generally make intertemporal decisions using much lower implicit discount rates than private individuals, already in the reference projection achieve a considerable amount of energy efficiency gains. In the Fragmentation case failure of price reform combined with a relative scarcity of capital, result in considerably higher consumption in most countries (with the exception of Israel which is less affected). In the MED-EU Initiatives strategy, the residential/commercial sectors are only slightly and indirectly (through somewhat accelerated price reform) affected since they are not specifically targeted. On the other hand, in these sectors the scope for specific energy consumption reductions is large as indicated in the results obtained for the Global Integration strategy (Table 8). It materialises primarily through the introduction of standards for lighting and appliances as well as insulation of buildings. The opposite trends are registered for the Fragmentation case where current account difficulties imply a slower turnover of energy consuming equipment. This is particularly the case in the residential sector due to a combination of subsidised electricity prices and low credit for the purchase of more efficient electrical equipment. In the MED-EU Initiatives strategy energy demand for transport is virtually unaffected. Even in the Global Integration efficiency gains are somewhat more limited than in other sectors, especially in countries like Morocco and Israel where prices in the reference projection are already high. In this context earlier retirement of vehicles is assumed and hence a faster turnover of the fleet with new vehicles generally having lower specific consumption characteristics than older vintages, but the process of transformation remains relatively slow. The proportion of new technologies like hybrid vehicles in the total car fleet increases from 8.1% in the reference to 12.8% in the Global Integration strategy in In general, using MENA-EDS, the transport sector as a whole and more specifically road transport appears to be less responsive to price signals than most other sectors. This is Fig. 1. Comparison of average fuel prices of the region in the alternative strategies.

8 8 P. Fragkos et al. / Energy Strategy Reviews xxx (2012) 1e12 Table 8 Changes in MED-9 final energy demand in illustrated by the relatively modest increase registered for the Fragmentation case despite the fact that in this context the average life of vehicles increases thus rendering the renovation of fleets and the subsequent improvements in efficiency slower. Hybrid vehicles shares remain so low that they merely represent a niche market. Overall the different projections examined have important implications in terms of the final energy intensity of GDP. Over the period 1990e2010 the MED-9 region was characterised by a complete stagnation in this index. The Reference projection represents a slight improvement ( 0.21% p.a. between 2010 and 2030). The Fragmentation case marks a deterioration (0.29% p.a.). In the MED-EU Initiatives strategy the improvement is of the order of 0.5% p.a., while for the Global Integration an improvement of 1.15% is registered and is directly comparable to the performance of EU-15 over the period 1990e2010. On the other hand, the EU-15 region is projected to improve this intensity at a rate of 1.9% p.a. in the period 2010e2030 [21] Impacts on primary energy supply In Mtoe Changes from reference in 2030 (in %) Present Reference MED-EU Global Fragmentation situation (2030) Initiatives Integration (2010) Final demand L5.5 L By sector Industry Residential Services Agriculture Transport By fuel Solids Oil Gas Electricity Traditional biomass Share of hybrids in car stock (in %) Hydrocarbon supply plays a major role in the economies of the region. For important exporters like Algeria and Libya hydrocarbon exports constitute a major item in their current account. On the other hand, predominantly importing economies are particularly vulnerable especially when international prices are high. Furthermore, there are countries like Egypt and Syria that have in the past produced exportable surpluses, are currently close to self-sufficiency but could become much more import dependent as fewer and smaller new fields are discovered while demand is growing at a fast pace. Under such conditions variations in the balance between demand and supply, such as those that result in the projections examined, can have important economic ramifications for the countries involved. Libya has good prospects for production expansion in the short term through enhanced recovery and in the longer term with the development of known fields and the parallel expansion of pipeline infrastructure (Table 9). Algerian production in the reference projection is projected to increase in the short term (2020) but decline in the longer term on the basis of the maturity of its oil fields. Similar trends but at a different scale apply to Egypt and Syria (the former is already marginally an oil importer, while the latter is projected to become one by 2030). The Global Integration strategy assumes an increase in FDI but not to the same extent as the MED-EU Initiatives. Consequently, primary production of oil increases compared to the reference projection but stands below the MED-EU Initiatives levels. On the other hand, the Global Integration strategy has a much wider scope for reduction in oil demand, especially because of the efficiency gains in the transport sector that play a key role in determining the overall demand reduction. The exportable surplus for Libya in 2030 goes from Mtoe in the reference projection to 167 Mtoe in the Global Integration strategy. The quantities for Algeria are 52.1 and 58.3 Mtoe respectively. However, even the higher figure for the Global Integration strategy does not prevent a reduction in Algerian oil exports between 2020 and In this context, Syria clearly remains in a net oil exporting position by 2030 (exports representing one third of primary production), while the import dependence of Egypt is only 22%. All other countries experience considerable reductions in net imports, compared to reference. According to the reference projection, gas production in MED-9 region is set for a major expansion (more than doubling between 2010 and 2030). This result is dominated by projected developments in Algeria and Egypt where increased production is sufficient to meet the rapidly expanding local demand and at the same time nearly double exportable surpluses (Table 10). Libya s exports also expand considerably albeit from a smaller base. Israeli production expansion from offshore fields in the Mediterranean produce a surplus equivalent to almost 50% of output. The remainder countries meet their expanding needs primarily from imports with intraregional trade projected to increase sharply from a very small base. The Global Integration strategy has pronounced impacts on gas demandesupply balances. Unlike the MED-EU Initiatives strategy, where the participation of Europe in a major decarbonisation effort means a substitution away from gas and hence a contraction of this crucial market for MED-9 suppliers, the Global Integration strategy sees an expansion in exports arising from higher production due to more investment in productive capacity as well as a sharp reduction in domestic demand due to the efficiency gains in final demand sectors and substitution towards renewables in the power generation sector. The exportable surplus of the region stands not only higher than the reference, but also a full 54% higher than the MED-EU Initiatives strategy. The negative investment climate assumed for the Fragmentation case produces pronounced effects on the supply of crude oil in the region. The effects (particularly strong in Libya) are due to delays in exploration programs that produce lower reserves and a more limited expansion of productive capacity. In this context Jordan does not undertake shale oil production by the end of the forecast period. The drop in production in conjunction with increased primary consumption of oil leads to major changes in hydrocarbon trade for the countries of the region. In the Fragmentation case net importers increase their dependence and net exporters reduce exported volumes as a consequence of higher domestic use. Viewing the MED-9 region as a whole, it is worth noting that in the reference projection it continues to be a net exporter of oil (131.3 Mtoe in 2030), while in the Fragmentation case the situation is reversed and it becomes a marginal net importer. Underinvestment in the hydrocarbon sector has a negative influence on gas productive capacity. Production in 2030 is 8.5% lower compared to the reference. Intraregional exchanges are either heavily reduced or completely eliminated as a matter of policy. Israel does not develop its LNG export capability and hence produces for the local market. Algeria registers a relatively small reduction in production and its primary consumption increases by more than 20%. Consequently its total exports drop from 93.6 Mtoe in the reference to 77.8 in the Fragmentation case. Egypt registers a bigger drop from 30.1 Mtoe to 11.4 Mtoe mainly as a consequence of the contraction of its export markets in Syria, Jordan and Lebanon.