WPS7697 Policy Research Working Paper 7697 Long-Term Energy Demand Forecasting in Romania Modeling Approach Sunil Malla Govinda R. Timilsina Development Research Group Environment and Energy Team June 2016 Policy Research Working Paper 7697 Abstract This study develops an end-use energy demand analysis would increase by 46 percent over the same period, because model for Romania to project energy demand by sector of increased household income and the expanded service and end-use for 2015–50. The study finds that Romania’s sector, which is relatively electricity intensive. Still, per energy demand in 2050 would be 34 percent higher than capita electricity consumption in Romania will be about the level in 2013. The industry sector would be the largest half the European Union 28 average. At the end-use level, final energy-consuming sector, surpassing the residential thermal processes in the industry sector, space heating in sector from 2025 onward. The services sector would exhibit the residential and services sectors, and road transporta- the fastest growth of energy consumption in line with the tion in the transport sector would be dominant throughout expected structural change from manufacturing to services. the study period. The study also shows that improve- Although population in the country is projected to drop by ment of energy efficiency in the heating system would be 7 percent in 2050 from the 2013 level, electricity demand the main channel to cut energy demand in the country. This paper is a product of the Environment and Energy Team, Development Research Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at gtimilsina@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Long-Term Energy Demand Forecasting in Romania: An End-Use Demand Modeling Approach Sunil Malla and Govinda R. Timilsina* Key words: Energy demand, end-use modeling, demand forecasting, climate change, Romania JEL Classification: Q41, Q47 _________________________ # This study builds on the work carried out under the Advisory Services Agreement on Romania Climate Change and Low Carbon Green Growth Program signed between the Ministry of Environment and Climate Change and the World Bank on July 23, 2013. The authors would like to thank Erika Jorgenson, Feng Liu, Jian Xie, Kulsum Ahmad, Mike Toman, Morgan Bazilian and participants of several workshops organized in Bucharest in 2013-2015 period for their valuable comments and suggestions. The authors would also like to thank country experts from Ministry of Environment Waters and Forest (MEMW) and Ministry of Energy (MOE) of Romania, for their constructive comments and invaluable insights in developing EEDA model for Romania.  We acknowledge the financial support from the European Union, Romanian Government and the World Bank. * Malla is a short-term consultant and Timilsina is the Senior Research Economist, Development Research Group, The World Bank, Washington, D.C. Timilsina (gtmilsina@worldbank.org) is the corresponding author. 1. Introduction Romania, with a population of 20 million, is the eleventh largest economy in the 28 member states of the European Union (EU-28). While the country, like other Eastern European economies, experienced economic downturn in the early 1990s during the political and economic transition, it observed an economic recovery soon in 1993. The entry into the EU in 2007 further accelerated growth, although it was badly hit by the 2009-10 financial crisis. Between 1995 and 2013, Romania’s per capita income increased by 73%, more than twice as fast as Germany (28%) and the EU-28 average (30%). Over the past decade, the country’s real gross domestic product (GDP) and disposable household income per capita grew on average by 3.5% and 7.2% per year respectively, well above the rate of the EU-28 as a whole (Eurostat, 2015a).1 Despite the impressive economic growth, the current levels of per capita GDP and energy consumption in Romania fall much below that of most other countries in the EU. As the Romanian economy expands and households’ incomes rise, more energy would be needed to satisfy the growing demand. The average monthly household expenditure on non-food goods, which currently represents 21% of the total expenditure, has increased by twofold (8% per year) between 2005 and 2013 (INS, 2015a). The household ownership of refrigerators grew on average by 5% per year, followed by vacuum cleaners (2.8%), washing machines (2.2%), televisions (1.8%), and passenger cars (1%) during the 2008-2013 period (INS, 2015b). If no action is taken, increasing energy demand will pose a challenge to Romania to comply with its greenhouse gas (GHG) obligations set forth by the EU with its member states.                                                              1 Despite a major slump in economic growth in 2009 (−7.1%) and 2010 (−0.8%), due to global economic and financial crisis, Romania’s real GDP grew on average by 7% per year during 2003−2008 and by 2% per year during 2011−2014. 2    The EU has set a number of short- and long- term quantitative climate and energy targets for its member countries, notably 20-20-20 climate and energy targets, 2030 framework for climate and energy policies and roadmap for moving to a low carbon economy in 2050. As part of the broader EU energy and climate strategies and policies, Romania is committed to energy security, energy efficiency and competitive energy market improvement, renewable energy promotion and low carbon green growth development. Further, based on the assessment of 2014 national reform and convergence programs for Romania, EU commission recommended energy program that focuses on improving efficiency of industries, thermal insulation of buildings and the rehabilitation of district heating systems (EU, 2014a). There exists only a limited number of empirical studies that examine future energy demand in Romania. A recent report (World Bank, 2014) expects a significant increase in final energy demand (FED), particularly in transport and services sectors. Using econometric techniques, Bianco et al. (2010) estimate long-run GDP and price elasticities of non-residential electricity consumption in Romania and finds that these elasticity values are quite low, i.e., 0.496 (income elasticity) and −0.274 (price elasticity). The sixth National Communication of Romania to the United Nations Framework Convention on Climate Change (MECC, 2013) projects that the country’s FED would increase by 28% between 2010 and 2030, and that residential and industry sectors are responsible for over 60% of this increase. Likewise, using PRIMES model, EU (2014b) projects country’s FED to grow by 21% between 2010 and 2050, with three-fourth of this increase is coming from industry and transport sectors. Analyzing historical FED over the past decade, ICEMENERG and ANRE (2012) find that country’s demand for energy in industry sector has declined, mainly due to shifting away from energy-intensive manufacturing industries, while it has increased in the rest of the sectors, thereby confirming a structural change of the economy. However, these studies lack analysis of 3    detailed long-term projections of energy demand at sub-sector and end-use levels Understanding of these detailed energy demand evolution is critical in implementing EU recommended energy and climate strategies in the country. In this study, we develop an end-use bottom-up model to examine the long-term energy demand at sub-sector level for Romania. 2. Current Energy Demand Structure in Romania Romania’s total FED was 21.8 Mtoe in 2013. Since 2009, the residential sector surpassed the industry sector as the largest consumer of final energy. The residential sector accounted for 36% of total FED in 2013, reflecting high heating demands in cold climate and an ageing housing stock. The country’s share of residential sector in total FED in the same year ranked the highest among EU-28 countries (Eurostat, 2015b). The industry sector including agriculture (31%), transport sector (24%) and the services sector (9%) are the second, third and fourth largest consumers of final energy. In Romania, FED has been in downward trend over the past decade, falling by 12% between 2004 and 2013. This is mainly due to structural changes and declining energy intensity in the industry sector. By fuel types, oil products accounted for 30% of total FED, with more than 76% consumed by the transport sector. Natural gas, mainly used in industry and residential sectors, accounted for 27%, while biomass and renewable wastes accounted for 17%, with more than 85% used in residential sector. Likewise, electricity accounted for 16%, followed by district heat (6%) and coal (3%) in total FED in 2013(EU, 2015). In per capita terms, country’s demand for electricity and energy are still much lower as compared to the respective average values for EU-28. For example, in 2013, Romanians consumed about one-third of electricity to that of Germans and about 37% to that of EU-28 average (Table 1). Likewise, country’s per capita FED (1,087 kgoe/capita) is less than half of 4    Germany and EU-28 average in the same year. In addition, the number of passenger cars per thousand people in Romania was 243 in 2013, much lower compared to Germany (543) and the EU-28 average (491) (Table 1). As economy grows and income rises, the number of passenger cars and transport energy demand is expected to grow in the foreseeable future. 900 500 800 450 Final energy intensity (kgoe/1000 US$) Transport−Germany Final energy consumption/capita (kgoe) 700 400 Transport−EU−28 Residential−Germany 350 600 Residential−EU−28 300 500 Residential−Romania 250 400 Industry−Romania 200 300 150 Transport−Romania Industry−EU−28 200 100 Industry−Germany 100 50 Services−EU−28 Services−Germany Services−Romania 0 0 0 5000 10000 15000 20000 25000 30000 35000 40000 0 5000 10000 15000 20000 25000 30000 35000 40000 GDP/capita (constant 2005 US$) GDP/capita (constant 2005 US$) Fig. 1. Relation between income and final energy demand by sector in per capita term (left), and income per capita and final energy intensity by sector (right) in Romania, Germany and EU-28 during 1992-2012 period. Source: World Bank (2015) and Eurostat (2015b). There is strong correlation between the levels of sectoral FED and income. For instance, in per capita terms, FED of both residential and transport sectors and income in Romania is much lower compared to Germany and the EU-28 average during 1992-2012 period (Figure 1, left). Similarly, there is a close correlation between income and sectoral final energy intensity. In particular, final energy intensity of industry sector relative to GDP per capita is declining in Romania, Germany and the EU-28 average during 1992-2012 period (Figure 1, right). In contrast, correlation between energy intensity of services sector relative to GDP per capita is about the same during the same period. However, these sectoral intensities are relatively much higher in Romania when compared with high-income countries like Germany and the EU-28 average. More specifically, average final energy intensity of industry in the country declined significantly by 70% during the same period. 5    Table 1 Overview of selected historical socio-economic and final energy consumption indicators in Romania, Germany and EU-28 Romania Germany EU-28 1992 2013 % chg. 1992 2013 % chg. 1992 2013 % chg. 92-13 92-13 92-13 GDP per capita (PPP, constant 2011 US$) 10,366a 18,182 75 32,919 a 43,444 32 25,603 a 36,925 44 Urban population (%of total) 54 54 0 73 75 3 71 76 7 Final electricity consumption/capita (kWh) 1,818 2,033 12 5,595 6,310 13 4,527 5,450 20 Final energy consumption (FED)/capita (kgoe) 1,192 1,087 -9 2,753 2,649 -4 2,221 2,179 -2 Industry, FED (% of total) 52 29 -44 29 28 -2 31 25 -19 Transport, FED (% of total) 15 24 66 28 29 3 28 32 13 Residential, FED (% of total) 23 35 54 28 27 -2 26 27 1 Services, FED (% of total) 2 8 255 13 16 22 11 14 30 Others, FED (% of total) 8 3 -62 2 <1 -97 4 3 -34 Final energy intensity (toe/million 2010 €) 629 243 -61 165 121 -27 170 129 -24 Import dependency (%) 30 19 -37 55 63 15 46 53 16 Passenger cars per 1000 people 70 235 236 480 543 13 361 491 36 Mean consumption expenditure per HH (€) 1,373b 5,514 c 302 25,228 b 29,330 c 16 25,010 c Note: a Data from 1995; b data from 1999; c data from 2010. Source: World Bank (2015) and Eurostat (2015b, c, d). 6    3. Methodology Normally two types of techniques are used for energy demand forecasting: an econometric approach, or an end-use accounting approach. The first approach is often used at the aggregated level such as total energy demand. In this approach the statistical relationship between energy consumption and macroeconomic variables, such as GDP is established based on historical data and the same relationship is used to forecast future energy demand. Such an approach is not applicable when detailed energy demand forecasting is needed at the end-use level because long time series of historical data on detailed end-use energy consumption are not available. An end- use accounting model, which does not need time series data, but relies on detailed data for a reference or base year, is normally employed for forecasting end-use energy demand at various sectoral levels.2 For this study, we also developed an end-use energy forecasting approach for Romania. For a given end-use in a given sector, the main elements of energy demand in our model are activity, structure and intensity. The drivers of future energy demand are the scenarios. Four energy consuming sectors (residential, services, industry and transport) are considered in the model. The services sector is further divided into 7 sub-sectors by type of buildings: office, educational building, hospital, hotel and restaurant, sport facilities, wholesale and retail store, and others (not classified elsewhere). Industry is divided into 4 sub-sectors: agriculture, construction, mining and quarrying, and manufacturing. The manufacturing sub-sector is further sub divided into 10 manufacturing industry types based on economic activities in the European Community classification (EC, 2008). Transport is further sub-divided into 5 sub-sectors (road, rail, air, inland waterways, maritime and pipeline) based on mode of transportation. The residential sector could be divided between rural and urban types but we did not know due to                                                              2 For more discussion on energy demand modeling, please refer to Bhattacharyya and Timilsina (2010). 7    lack of data. The sector, sub-sector and end-use classifications used in the model are presented in Figure 2. Residential Transportation Industry Services • Cooking • Cooking • Agriculture • Road • Lighting • Lighting • Mining & • Passenger • Water heating • Water heating quarrying • Freight • Space heating • Space heating • Construction • Railways • Electric • Space cooling • Manufacturing • Passenger appliances • Iron & steel • Freight • Electric • Refrigerator appliances • Non−ferrous metals • Air • Air conditioner • Refrigerator • Chemical & • Passenger • Fan petrochemical • Freight • Air conditioner • Washing machine • Non−metallic • Waterways • Fan • Dish washer minerals • Passenger • Washing machine • Vaccum cleaner • Food & tobacco • Freight • Dish washer • Textile & leather • Microwave oven • Vaccum cleaner • Pipelines • TV • Paper, pulp & print • Microwave oven • Freight • Stereo • Transport equipment • TV • Computer • Machinery • Computer • Wood & wood • Iron • Other appliances products • Other appliances • Other manufacturing Fig. 2. Simplified classification of sector, sub-sector and end-use classification in Romania. The energy demand for each end-use categories is driven by one or several demographic, socio-economic and technological parameters, whose values are given as part of the scenarios. Demand of energy use by each sub-sector and by end-use is calculated using the following general equations: E X (t )   E x , X ( t )   E y , x , X (t ) , x  X (1) x y E y , x (t )  A y , x (t )  S y , x (t )  E I y , x (t ) (2) where EX is the total energy demand for sector X, A is the activity level, S is the structure and EI is the energy intensity. The small cap subscripts x and y represent sub-sector and end-use, respectively. For example, residential sector end-uses include space heating, water heating, air- 8    conditioning, cooking, lighting and use of electric appliances. Activity level may be value-added or domestic production for industry sector, domestic demand and commercial floor area for services sector, population, living floor area and ownership of electrical appliances for residential sector, and passenger kilometer (pkm) or ton-kilometer (tkm) for transport sector. Likewise, sectoral structure is the mix of activities within a sector and sub-sectors, such as the share of domestic production in manufacturing industries, and energy intensity is the energy use per unit of activity, such as energy use per domestic production for manufacturing industries. Depending on data availability, breakdown of energy demand by sub-sectors and end-uses for each sector are different. Note that demand for energy is not disaggregated by fuel types. The fuel-mix largely depends on the technological possibilities of supply and their relative prices, which are outside the scope of this paper. However, due to its non-substitutability nature and importance, demand for electricity by sector is calculated separately in FED. In addition to estimating transport energy demand, activity-structure-intensity-fuel (ASIF) framework approach is used for calculating travel demand. The following equations are used: TDk ,T (t )   TD j , k ,T (t )   TDD j ,k ,T (t )  TI j , k ,T (t ) , j T (3) j j TI j ,k ,T (t )  TI j ,k ,T (t  n)  1  R j ,k ,T (t )  n (4) R j , k ,T (t )  TE j , k ,T (t )  PE j , k ,T (t )  RE j , k ,T (t ) (5) where TD is travel demand, TDD is transport domestic demand (activity variable), TI is transport mobility intensity, R is growth rate of overall change in mobility intensity, TE is growth rate of travel efficiency improvement change, PE is growth rate of population change, RE is growth rate of rebound effect change and T is transport sector, t is time period and n is number of period. The small cap subscripts j and k represent transport mode (e.g., air, road, rail, water or pipeline) and activity type (passenger or freight), respectively. In this framework, travel demand is the function 9    of mode, mobility intensity, travel efficiency, population change and rebound effect. Passenger travel demand is measured in pkm or number of passengers, and freight travel demand is measured in tkm or ton. In Romania, land transport includes road, rail and pipeline, water transport includes maritime and inland waterways, and air transport includes aircrafts. Excluding pipeline, each of these transport modes are further sub-divided by activity (e.g., passenger and freight). Since road dominates Romania’s transportation system, it is further sub-divided by private and public road transport. 4. Data, Scenario Description and Key Assumptions Required data for the study are grouped under three categories: energy, socio-economic and demographic, and technological data. The primary source of these data are Romania’s National Institute of Statistics (INS), EU’s Eurostat, the World Bank’s World Development Indicators, International Energy Institute (IEA)’s energy balances, and Building Performance Institute (BPIE)’s buildings data. Since none of these publications provide complete dataset, additional dataset are compiled from Ministry of Environment and Climate Change (MMSC)’s sixth National Communication on Climate Change report (MECC, 2013), International Telecommunication Union (ITU)’s Measuring the Information Society 2014 report (ITU, 2014), ENTRANZE project report (Atanasiu et al., 2012) and EU’s JRC Scientific and Policy reports (Bertoldi et al., 2012; Pardo et al., 2012) , in particular, for calibrating base year and activity parameters used in projecting energy demand. The starting year of the analysis is 2013 (base year) and the projection is made through the year 2050 with 5 years of interval starting from 2015. Three scenarios are considered: baseline, low demand and high demand scenarios. These scenarios are differentiated primarily by their 10    underlying assumptions about socio-economic and technological factors. The baseline scenario takes into account the current trends on socio-economic development, sectoral energy-use patterns and technological progress. It reflects the path of future energy demand given a continuation of current trends and policies. To illustrate increase in end-use energy demand (optimistic pathway) with respect to the baseline scenario, a high scenario is constructed. This scenario is characterized by the Romanian economy and population growing faster than the baseline scenario. The high scenario also reflects households with high home electric appliances and private vehicle ownership, and more passenger and freight transport mobility. Overall, in this scenario, the general development mode of Romanian economy is optimistic. In contrast, to illustrate decrease in end-use energy demand (pessimistic pathway) with respect to baseline scenario, a low scenario is constructed. In this scenario, economic and population growth are slightly lower compared to baseline scenario. Due to slower economic growth, domestic production and domestic demand are also lower. The selected key driving variables under the baseline and two alternative scenarios are summarized in Table 2. 11    Table 2 Key driving variables under the baseline, high demand and low demand scenarios for selected years (Index 2013=100, unless otherwise stated) Base case High Low 2020 2030 2050 2020 2030 2050 2020 2030 2050 Population a 99 95 90 99 96 91 98 95 87 GDP 127 163 233 133 171 244 122 156 223 Residential effective floor space 103 107 113 105 112 124 102 105 108 Appliances ownership Refrigerator/freezer 108 118 140 111 128 160 104 109 120 Washing machine 105 112 126 105 112 126 105 112 126 Television 107 118 138 107 118 138 107 118 138 Computer 142 204 327 142 204 327 142 204 327 Others 112 129 163 112 129 163 112 129 163 No. of residential dwellings with AC 135 189 285 135 190 290 135 187 275 Services buildings floor space 119 143 192 124 156 232 115 131 160 Offices (public and private) 113 132 163 117 144 197 108 120 135 Educational buildings 113 132 163 117 144 197 108 120 135 Hospitals 126 155 227 131 169 273 122 142 188 Hotels and restaurants 113 132 163 117 144 197 108 120 135 Sport facilities 126 155 227 131 169 273 122 142 188 Wholesale and retail trade 119 143 192 124 156 232 115 131 160 Others (n.e.c.) Services buildings demolition rate (%) 1.3 2.3 1.0 1.3 2.3 1.0 1.3 2.3 1.0 Services buildings occupancy rate (%) 85 85 85 85 85 85 85 85 85 Industrial domestic production Agriculture 107 119 130 111 129 156 103 109 109 Manufacturing Iron and steel 98 104 107 99 110 125 92 93 89 Chemical and petrochemical 127 170 279 140 200 371 131 170 267 Non-metallic minerals 115 136 162 115 142 188 107 120 133 Food and tobacco 122 143 189 128 158 231 120 134 165 Textile and leather 122 143 189 128 158 231 120 134 165 Paper, pulp and print 100 107 122 102 115 145 95 97 105 Transport equipment 134 196 311 139 212 371 130 180 261 Machinery 136 204 330 140 221 394 131 187 278 Wood and wood products 122 143 189 128 158 231 120 134 165 Other manufacturing (n.e.c.) 111 135 180 115 140 208 107 118 145 Mining and quarrying 102 111 129 105 119 147 98 101 102 Construction 106 152 235 108 162 278 101 137 197 Transport domestic demand Transport (Air) 109 127 201 115 134 211 105 122 193 Transport (Road-Freight) 143 193 286 150 203 301 137 185 275 Transport (Road -Passenger) 112 142 222 118 149 234 108 136 214 Transport (Rail -Freight) 102 121 194 107 127 204 98 116 186 Transport (Rail -Passenger) 116 147 218 122 154 229 111 141 210 Transport (Water - Freight) 97 111 194 102 116 204 93 106 187 Transport (Water - Passenger) 98 113 208 103 119 218 94 108 199 Gas pipeline 112 126 142 117 132 149 107 121 136 Rebound effect 103 107 115 103 107 115 103 107 115 Note: a Based on Eurostat’s main variant (baseline), high life expectancy variant (high) and low fertility variant (low) population projection. 12    5. Results and Discussion 5.1. National level final energy demand In this section, results of FED by sector are discussed under three scenarios. Note that FED is presented in two broad categories: electric and non-electric. In the baseline scenario, demand for electricity is projected to increase by an annual average of 1.3% (Table 3). Industry remains the single largest electricity user, accounting for almost half of total electricity demand over the projected period. By 2050, quantity of electricity demand in services, one of the fastest growing sectors in the country, is projected to be about the same as that of residential sector. Despite transport’s share in total electricity demand is small (3.5%) in 2050, this sector exhibits the fastest rate of expansion in percentage terms at an average rate of 2% per year, primarily due to increasing use of electric vehicles. Table 3 Electricity demand by sector in the baseline scenario, 2013-2050 (ktoe) 2013 2015 2020 2025 2030 2035 2040 2045 2050 Industry 1,689 1,703 1,869 2,060 2,234 2,393 2,570 2,721 2,867 Residential 1,023 1,035 1,086 1,132 1,170 1,211 1,258 1,303 1,325 Services 685 719 810 880 954 1,035 1,119 1,196 1,276 Transport 96 100 110 122 135 149 164 181 200 Total 3,493 3,557 3,875 4,194 4,493 4,788 5,111 5,401 5,668 However, demand for electricity is markedly different across the scenarios. In high scenario, which reflects optimistic socio-economic development in the country, electricity demand is projected to grow on average by 1.8% per year during 2013-2050, reaching almost 6,681 ktoe in 2050, an increase of 91% from 2013 value. Electricity demand expands much more rapidly in high scenario compared to baseline scenario to a level in 2050 that is 18% higher (Figure 3). In contrast, electricity demand expands much slower in low scenario, which reflects pessimistic 13    socio-economic development in the country, at an average rate of 0.9% per year to a level in 2050 that is 13% lower from the baseline. 7 High scenario Historical a Projected + 1013 ktoe (+18%) Final electricity demand (Mtoe) 6 Baseline scenario - 746 ktoe (- 13%) 5 Low scenario 4 3 2 1 0 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 Baseline High Low Fig. 3. Historical and projected final electricity demand by scenario in Romania, 1990-2050. Note: a historical data are taken from INS (2015c). 6500 EU-28 6000 Electricity consumption (kWh/capita) 5500 Germany 5000 4500 4000 3500 Romania 3000 2500 2000 1500 1000 2013 2015 2020 2025 2030 2035 2040 2045 2050   Fig. 4. Final electricity consumption per capita in the baseline scenario, 2013-2050. Note: Data for EU-28 and Germany are taken from Eurostat (2015b) and IEA (2013, 2014) 14    As a result of rising demand for electricity and declining population in the country over the projection period, per capita final electricity consumption is projected to increase. For example, in the baseline, it is projected to increase from 2,033 kWh in 2013 to 3,698 kWh in 2050, an increase of 82% (Figure 4). However, the differences in per capita electricity consumption in Romania and other advanced countries in the EU remain very large. In 2040, per capita electricity consumption in Romania amounts to 3,239 kWh, while it is 5,710 kWh in Germany and 6,294 kWh in EU-28 (average). Note that Romania’s projected per capita electricity consumption in 2050 is much lower than that of Germany and average EU-28 in 2013. As a reference, if electricity use by each Romanian increased to the 2013 level of German, Romania’s electricity use would increase by threefold. Table 4 Final non-electric energy demand by sector in the baseline scenario, 2013-2050 (ktoe) 2013 2015 2020 2025 2030 2035 2040 2045 2050 Industry 5,078 5,118 5,507 5,941 6,353 6,723 7,109 7,397 7,669 Residential 6,699 6,691 6,716 6,716 6,676 6,659 6,667 6,670 6,586 Services 1,101 1,164 1,316 1,427 1,541 1,677 1,818 1,946 2,077 Transport 5,182 5,279 5,531 5,740 5,957 6,199 6,452 6,620 6,793 Total 18,060 18,252 19,070 19,824 20,527 21,258 22,046 22,633 23,125 As economy grows, Romania’s final non-electric energy demand is also projected to increase. Between 2013 and 2050, total final non-electric energy demand is projected to increase on average by 0.7% per year, reaching 23,125 ktoe in 2050, in the baseline (Table 4). The share of industry in total non-electric energy demand, increases gradually, from 28% in 2013 to 33% in 2050. In contrast, the share of residential in the total non-electric energy demand is slightly declined over the projected period, at an average rate of 0.05% per year, mainly due to declining population. By 2035, industrial non-electric energy demand surpasses the residential sector 15    mainly due to strong growth in overall industrial production. Despite declining population and improvement in vehicle fuel-economy during the projected period, rising number of motor vehicles and increasing economic activities will lead to rise in demand for energy. Transport sector’s non-electric energy demand is projected to increase steadily, from 5,182 ktoe in 2013 to 6,793 ktoe in 2050, an increase of annual average of 0.73%. Although services sector’s share in total non-electric energy demand is small (13%) in 2050, this sector has the fastest rate of growth in percentage terms, at an average rate of 1.7% per year reflecting the structural change in the country from manufacturing to service sectors . 25 Change in thermal energy demand (%) 20 15 10 5 0 -5 -10 -15 -20 2020 2030 2050 2020 2030 2050 High Low Transport Residential Industry Services Fig. 5. Changes in thermal energy demand by sector in high and low scenario compared to baseline scenario. The demand for final non-electric energy and travel are quite different across the scenarios. In the high scenario, the changes in industry’s non-electric energy demand compared to the baseline scenario, are projected to increase at much higher percentage rate than in other energy consuming sectors. For example, between 2020 and 2050, non-electric energy demand in industry increases from 5% to 23%, while it increases from 4% to 20% in services, 2% to 8% in 16    residential and 1% to 7% in transport (Figure 5). In the low scenario, changes in the services sector’s energy demand is projected to decrease at much higher rate than in other sectors. For example, the thermal energy demand in services decreases from 4% in 2020 to 17% in 2050, while it decreases from 3% to 12% in industry, 1% to 9% in residential and 1% to 4% in transport over the same time period. Although comparing energy demand forecast between studies is not straightforward, due to different methodologies and assumptions, we have compared our study with EU’s energy, transport and GHG emissions trends to 2050 (EU, 2014b), Ministry of Environment and Climate Change’s sixth national commutation (MECC, 2013) and the National Commission for Prognosis (CNP)’s energy balance forecast (CNP, 2015) for Romania’s final energy demand. The comparison suggests some similarities in magnitude of final energy demand. For example, the EU estimate of total FED in 2050 is 27.3 Mtoe compared to our estimate of 28.8 Mtoe. Likewise, EU estimate of total final electricity demand in 2050 is 5.7 Mtoe compared to this study estimate of 5.8 Mtoe. The possible reason for slightly lower EU estimate of final energy and electricity demand compared to this study is the assumption of country’s lower economic growth in EU study. Further, the MECC estimate of total FED without any measure (28 Mtoe) in 2030, the latest year reported, is slightly higher compared to this study in the baseline scenario (25 Mtoe), possibly due to higher economic growth assumption. The CNP made projection of FED by sector up to year 2018. For comparison, the CNP’s FED estimate of residential (7.8 Mtoe), industry (7.1 Mtoe) and transport (5.5Mtoe) sectors in 2015 is similar to this study in magnitude. 17    5.2. Residential sector The residential sector is Romania’s largest energy consumer. Romanian households spend more than 13% of their income on energy, one of the highest rates in the EU (EU, 2014a). In 2013, this sector accounted for 35% of total final energy consumption, using natural gas, LPG and solid biomass as the major fuels. In the baseline, residential FED is projected to grow on average 0.1% per year from 2013 to 2050 to reach 7,911 ktoe (Table 5). The residential energy demand varies markedly in the alternative scenarios. For example, it is projected to increase from 7,722 ktoe in 2013 to 8,547 ktoe in 2050 in high scenario, at an average of 0.3% per year, it is projected to decrease (0.2% per year) in low scenario, to reach 7,202 ktoe in 2050. Non-electric energy demand, mainly used for space and water heating, contributes the most in total residential FED in all three scenarios. The contribution of electricity, mainly used for running electric appliances and lighting, is relatively small in total residential FED in all three scenarios. Table 5 Residential sector end-use energy demand by scenario, 2013-2050 (ktoe) Baseline High Low 2013 2020 2030 2050 2020 2030 2050 2020 2030 2050 Space heating a 4,638 4,707 4,777 4,861 4,814 5,023 5,380 4,624 4,551 4,310 Air conditioning 31 42 59 89 42 59 90 42 58 86 Water heating 1,299 1,267 1,201 1,094 1,269 1,207 1,113 1,266 1,193 1,060 Cooking 790 772 731 667 772 735 679 771 727 646 Lighting 218 223 231 244 228 243 270 219 220 217 Electric appliances 746 791 849 955 801 873 1014 782 823 884 Refrigerator/freezer 240 248 256 279 257 278 325 239 235 230 Washing machine 118 120 122 127 120 123 129 120 122 122 Television 176 184 194 210 184 195 214 184 193 202 Computer 28 39 53 79 39 54 81 39 53 76 Other appliances 184 200 223 261 201 224 266 200 221 252 Total 7,722 7,802 7,847 7,911 7,926 8,140 8,547 7,704 7,573 7,202 a Note: climate corrected. 18    At the end-use level, space heating (climate corrected) is by far the largest energy consumer across all scenarios. In the baseline, space heating accounted for 60% of total residential FED in 2013 and it is projected to remain about the same during 2015-2050 period (Figure 6). In contrast, energy demand for water heating and cooking is projected to decline slightly both in absolute and percentage terms during 2015-2050 period. For example, energy demand for these two end-uses is projected to decline on average 0.5% per year in the baseline (Table 5). This is mainly due to the country’s declining population and improvement in efficiency of heating and cooking devices during the projected period. It follows the similar trend in alternative scenarios, increasing in magnitude in the high scenario and decreasing in magnitude in the low scenario compared to the baseline scenario. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2013 2015 2020 2025 2030 2035 2040 2045 2050 Space heating Air conditioning Water heating Cooking Lighting Electric appliances   Fig. 6. Percentage share of end-uses in total residential final energy demand in the baseline scenario. Space heating, followed by water heating and cooking, are the major end-use services in the residential sector. The highest share of energy requirements for space heating in the country is mainly due to long cold winter with higher heating degree days. Apart from outdoor temperature, 19    many other factors influence energy demand for space heating, including the size and type of dwellings, and the efficiency of the heating system and equipment.3 Space heating therefore represents the largest opportunity to reduce residential energy demand, for example, by using more efficient heating equipment and by changing energy-mix. For space heating, natural gas and derived heat are commonly used in urban households, while solid biomass fuels including charcoal is commonly used in rural households. In 2013, about half of total energy used for residential space heating came from solid biomass, followed by natural gas (31%) and derived heat (18%) (Eurostat, 2015b). In the same year (2013), 80% of total residential energy is used for heating (space and water heating), 7% is used for cooking and the remaining 13% is used lighting and electric appliances including air conditioning. As the economy grows and income rises, the use of household electrical appliances and corresponding demand for electricity are also projected to increase. For example, demand for electricity by electric appliances is projected to increase from 746 ktoe in 2013 to 955 ktoe in 2050, at an average rate of 0.7% per year in the baseline. In particular, residential use of refrigerators, washing machines and televisions combined accounts for 72% of electricity demand by electric appliances during the 2015-2050 period. Electricity demand for lighting is projected to increase at a slightly lower rate of 0.3% per year to reach 244 ktoe in 2050. Despite air conditioning’s share in total residential energy demand is small (1%) in 2050, this residential end-use has the fastest rate of growth in percentage terms, at an average rate of 2.9% per year in the baseline scenario.                                                              3 Despite declining population during the 2013-2050 period, demand for heating energy is projected to increase mainly due to increase in size of living floor space in the country. For example, between 2013 and 2050, living floor space is projected to increase on average by 0.3% per year. 20    5.3. Services sector The importance of the services sector for Romania’s energy policy has grown significantly over the past decade. In 2013, services sector accounted for more than half of country’s GDP and about 8% of total FED (INS, 2015c, d). Since 2002, energy demand has increased by 41% as a result of an average growth of 3.5% per year. This sector is also the most heterogeneous sector of the economy that includes wide range of energy consumers. Demand for energy by services is projected to grow at much higher rate than other energy consuming sectors in all three scenarios. For example, services total FED is projected to double from 1,785 ktoe in 2013 to 3,353 ktoe in 2050, at an average rate of 1.7% per year in the baseline (Table 6). Relative to the baseline scenario, it follows higher trends in the high scenario and lower trends in the low scenario. Table 6 Services sub-sector end-use energy demand by scenario, 2013-2050 (ktoe) Baseline High Low 2013 2020 2030 2050 2020 2030 2050 2020 2030 2050 Space heating 840 1,006 1,179 1,594 1,046 1,287 1,919 968 1,080 1,322 Space cooling 94 113 135 188 117 147 227 108 123 156 Water heating 149 178 208 280 185 227 338 171 191 233 Lighting 231 273 326 437 284 356 527 263 299 363 Others 473 556 647 853 578 706 1,027 535 592 707 Total 1,785 2,126 2,495 3,353 2,210 2,724 4,038 2,045 2,285 2,781 Note: b Others include energy demand for cooking (mainly in hotel/restaurant) and electricity for appliances and lighting in public places. At the end-use level, similar to the residential sector, services sector space heating is projected to account for most of the energy demand. In 2050, the share of space heating in total services FED accounts for 48% in the baseline, followed by others (25%), lighting (13), water heating (8%) and air cooling (6%). At the sub-sector level, excluding energy demand for others, whole sale and retail store, and office buildings combined accounts for more than half of services FED in all three scenarios during the projected period (Figure 7). Hotel and restaurant, hospital, 21    educational building, sports facility and n.e.c. (not elsewhere classified) sub-sectors account for remaining half of services energy demand in descending order during the projected period. 3500 3000 2500 2000 ktoe 1500 1000 500 0 Base High Low Base High Low Base High Low 2013 2020 2030 2050 Office Educational Hospital Hotel Sport facilities Wholesale n.e.c. a Fig. 7. Services sub-sector energy demand by scenario, 2013-2050. Note: a n.e.c. include institutional buildings, warehouses and other non-specified service industries. Lighting Water heating Space cooling Space heating 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Office Educational Hospital Hotel/restaurant Sport facility Wholesale n.e.c.   Fig. 8. Services end-use and sub-sector energy demand in the baseline scenario, 2050. Likewise, in 2050, both at the end-use and sub-sector levels, the share of space heating is projected to account for the highest (37%) in wholesale and retail stores to the lowest (4%) in the 22    n.e.c. sub-sector (Figure 8). In the case of lighting, it accounts for the highest (24%) in hospitals to the lowest (4%) in the sport facility sub-sector in the baseline. This is followed by water heating, in the range of 4% in n.e.c. to 24% in hotel and restaurant, and space cooling, in the range of 3% in n.e.c. to 29% in offices in 2050. 5.4. Industry sector Despite a sharp decline in industry’s total FED since 1990, Romania remains heavily dependent on energy-intensive manufacturing industries. In 2013, about 87% of industry’s total FED is consumed by manufacturing industries. In the same year, close to 70% of total manufacturing energy demand is consumed by a handful of energy intensive industries, such as iron and steel, chemical and petrochemical, and non-metallic minerals (Eurostat, 2015b). In terms of energy types, coke, natural gas and electricity are mainly used in iron and steel industry, while oven coke, while natural gas, electricity and refinery gas are mainly used in chemical and petrochemical industries, and petroleum coke, natural gas and electricity are mainly used in non- metallic minerals industries. The industry’s total FED is projected to increase from 6,767 ktoe in 2013 to 10,536 ktoe in 2050, at an average rate of 1.2% per year, in the baseline (Table 7). Demand for electricity in this sector is projected to increase at much higher rate (1.4% per year) than the demand for non- electric energy (1.1% per year) in the baseline during 2015-2050 period. Note that electricity is used for only electric motors and others, while non-electric (thermal) energy is used for motive, thermal-electric and heat requirements. Relative to baseline scenario, the demand for industrial energy is projected to increase in the high scenario and decrease in the low scenario. 23    Table 7 Industry sub-sector and end-use final energy demand by scenario, 2013-2050 (ktoe) Baseline High Low 2013 2020 2030 2050 2020 2030 2050 2020 2030 2050 Electric motor 1,013 1,121 1,341 1,720 1,165 1,465 2,094 1,088 1,241 1,487 Agriculture 42 45 48 51 46 52 61 43 44 43 Manufacturing 932 1,036 1,240 1,596 1,078 1,357 1,947 1,006 1,149 1,383 Mining and quarrying 12 12 13 14 12 14 16 12 12 11 Construction 27 28 40 59 29 42 70 27 36 49 Others (electricity) 676 748 894 1,147 777 977 1,396 725 827 991 Agriculture 28 30 32 34 31 35 41 29 29 28 Manufacturing 621 691 827 1,064 718 904 1,298 671 766 922 Mining and quarrying 8 8 9 10 8 9 11 8 8 8 Construction 18 19 26 39 19 28 46 18 24 33 Motive/thermo-electric 537 577 707 916 592 761 1,088 553 644 767 Agriculture 253 268 294 316 277 319 378 258 269 265 Manufacturing 71 85 104 139 88 113 167 82 95 119 Mining and quarrying 20 20 22 25 21 24 29 20 20 20 Construction 193 203 286 436 207 307 514 193 260 364 Thermal energy 4,540 4,817 5,382 6,163 5,029 5951 7633 4,696 5,038 5,440 Agriculture 131 132 136 130 136 148 155 127 125 109 Manufacturing 4,251 4,634 5,286 6,303 4,857 5,867 7,833 4,537 4,969 5,586 Mining and quarrying 3 3 3 3 3 3 4 3 3 3 Construction 156 161 220 316 164 236 373 153 200 264 Total (industry) 6,767 7,376 8,587 10,536 7,695 9,457 12,942 7,185 8,009 9,206 At the end-use level, thermal (heat) energy use is projected to account for most of the energy demand in the industry sector during the projected period in all three scenarios. For example, in 2050, the share of thermal energy in total FED accounts for 64% in the baseline. This is followed by electric motor (16%), others (11%), and motive and thermal-electric power (9%) (Figure 9, right). At the sub-sector level, manufacturing industry is by far the largest end-user in terms of total industry energy use. In 2050, manufacturing industry is projected to account for 86% of total industry energy demand, followed by construction (8%), agriculture (5%) and mining and quarrying (1%) in the baseline (Figure 9, left). In manufacturing industries, iron and steel, chemical and petrochemical, and non-metallic minerals combined are projected to account for two-thirds of total manufacturing energy demand 24    through the projection period in the baseline (Figure 10). The other notable manufacture sub- sectors include machinery, and food and tobacco. The share of these two sub-sectors combined in total manufacturing energy demand is projected to increase from 15% in 2013 to 22% in 2050 in the baseline. In absolute values, the top three energy consuming manufacturing sub-sectors are chemical and petrochemical (3,072 ktoe), iron and steel (1,424 ktoe) machinery (1,052 ktoe) and non-metallic minerals (1,015) in 2050. 100 100 80 80 60 60 % % 40 40 20 20 0 0 2013 2020 2030 2050 2013 2020 2030 2050 Agriculture Manufacturing Electric motor Others a Mining and quarrying Construction Motive/thermal-electric Thermal (heat)     Fig. 9. Share of sub-sector (left) and end-use (right) industrial energy demand in the baseline scenario. Note: a Others include electricity for lighting, electric appliances, electro-chemical process, electro-thermal process and refrigeration. 2050 2030 2020 2013 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Iron & steel Chemical & petrochemical Non-metallic minerals Food & tobacco Textile & leather Paper, pulp & print Transport equipment Machinery Wood & wood products Others (n.e.c.)   Fig. 10. Industrial manufacturing sub-sector final energy demand in the baseline scenario, 2013-2050. 25    8000 7000 6000 5000 ktoe 4000 3000 2000 1000 0 2020 2030 2050 2020 2030 2050 2020 2030 2050 2013 Base High Low Furnace and direct heat Steam generation Space and water heating Fig. 11. Industrial manufacturing thermal energy demand by scenario, 2013-2050.  In the industry sector, three temperature intervals are considered to describe the quality of thermal energy demanded by the manufacturing industries. The low temperature (less than 100 °C) corresponds to processes as washing, rinsing, and water and space heating of the industrial facilities, the medium temperature (100 °C to 400 °C) corresponds to steam generation, and the high temperature (more than 400 °C) corresponds to furnace and direct heat. The thermal energy use for processes such as washing, rinsing and space heating in agriculture, mining and quarrying, and construction sub-sectors, is relatively small. In 2050, the share of these three sub- sectors combined in total thermal energy demand is projected to be only 6.6%, about 449 ktoe in the baseline (Table 7). In the case of manufacturing industries, the demand for furnace and direct heat (high temperature) is projected increase from 2394 ktoe in 2013 to 2948 ktoe in 2050, an increase of 0.6% per year (Figure 11). However, its share in total thermal energy demand is projected to decrease from 56% in 2013 to 47% in 2050, mainly due to increase in the share of space and water heating (low temperature) in total thermal energy demand in the baseline. 26    5.5. Transport sector Improvement of transportation remains very high on the policy agenda in Romania. In 2013, Romania’s transport energy demand is 5,278 ktoe or 24% of total FED. Despite decrease in country’s total FED, energy use in transport has increased steadily (1.3% per year) over the past decade, mainly due to a surge in passenger and freight road traffic. In 2013, 90% of Romania’s energy demand in the transport sector comes from road transportation and this share remains about the same over the projected period in all scenarios. The remaining transport energy demand is consumed mainly in air and rail transportation. Despite vehicle technology and fuel- economy improvements, transport energy demand is projected to increase during 2013-2050 period. Between 2013 and 2050, it is projected to increase by an average 0.8% per year to reach 6,993 ktoe in 2050 in the baseline, compared to on average 1% and 0.7% per year in high and low scenario, respectively (Table 8). Unless significant policy measures are adopted in promoting increasing use of mass transit system such as railways, road will remain the main transportation mode in the foreseeable future. As a consequence, this will put pressure on demand for oil and associated negative air pollution impact. Table 8 Transport sub-sector final energy demand by scenario, 2013-2050 (ktoe) Baseline High Low 2013 2020 2030 2050 2020 2030 2050 2020 2030 2050 Air 211 215 232 313 218 240 336 214 228 301 Road 4,766 5,118 5,508 6,211 5,189 5,696 6,684 5,082 5,416 5,987 Rail 255 263 304 393 267 314 423 262 299 379 Inland waterways 42 41 44 71 41 46 77 40 43 69 Pipeline 4 4 4 6 4 4 6 4 4 5 Total (transport) 5,278 5,641 6,092 6,993 5,720 6,301 7,526 5,602 5,990 6,741 Over the past decade, demand for passenger travel has been steadily increasing, while demand for freight travel has been declined slightly in the country. Since 2004, total passenger 27    movement from road, rail and inland waterways combined has increased on average by 2.6% per year, while freight movement has decreased on average by 0.3% per year. The decline in country’s freight movement over the past decade is mainly due to slump in economic growth in 2009 and 2010. However, as economy grows and the number of motorized vehicles increases, the passenger and freight travel demand is projected to increase over the 2013-2050 period. For example, excluding air and maritime transport, the passenger travel demand is projected to increase on average by 1.4% per year, reaching 174 Gpkm in 2050, and the freight travel demand is projected to increase on average by 1.3% per year, reaching 98 Gtkm in 2050 in the baseline (Table 9). Compared to baseline scenario, it follows the similar trends with higher travel demand in the high scenario and with lower travel demand in the low scenario. Table 9 Travel demand by transport mode by scenario, 2013-2050 (Gpkm, Gtkm) a Baseline High Low 2013 2020 2030 2050 2020 2030 2050 2020 2030 2050 Road passenger public 20.7 21.6 23.6 30.8 22.7 24.8 32.3 20.7 22.7 29.5 passenger private 80.4 82.4 89.9 135.9 86.5 94.4 142.7 79.1 86.3 130.4 freight 34.0 38.8 44.1 57.5 40.7 46.3 60.3 37.2 42.4 55.2 Rail passenger 4.4 4.8 5.4 6.9 5.1 5.7 7.2 4.6 5.2 6.6 freight 12.9 12.4 13.6 18.4 13.0 14.3 19.3 11.9 13.0 17.6 Waterways passenger 0.02 0.02 0.02 0.02 0.02 0.02 0.03 0.01 0.02 0.02 freight 12.2 11.6 12.9 21.5 12.2 13.6 22.6 11.1 12.4 20.6 Pipeline freight 0.8 0.9 1.0 1.0 0.9 1.0 1.1 0.9 0.9 1.0 Air passenger 10.7 11.1 11.7 16.3 11.6 12.3 17.1 10.6 11.2 15.7 freight 0.03 0.03 0.04 0.05 0.04 0.04 0.06 0.03 0.04 0.05 Maritime passenger 0.05 0.06 0.07 0.08 0.06 0.07 0.09 0.06 0.06 0.08 freight 43.6 52.0 62.7 83.6 54.6 65.8 87.8 49.9 60.2 80.3 Note: a Passenger and freight travel demand for air and maritime are expressed in million passenger and million ton. In Romania, road transport dominates both passenger and freight travel demand in the past and this trend is projected to continue during the projected period. For example, the road passenger travel demand is projected to grow on average by 1.4% per year, to reach 167 Gpkm 28    in 2050 from 101 Gpkm in 2013 in the baseline (Table 9). More specifically, private passenger travel demand dominates road transport. Roughly 80% of road passenger travel demand is projected to be met by private passenger vehicles (mainly cars) in the baseline (Figure 12). The remaining 20% of total road passenger travel demand is projected to be met by public passenger vehicles (mainly buses and mini-buses). The share of railways and inland waterways in total passenger travel demand is relatively small compared to passenger travel by road transport in the country. In 2050, the share of railways and waterways combined accounts to only 4% of total passenger travel demand in the baseline scenario. In addition to lack of railway infrastructure, the projected decline in population and rising incomes over the study period will most likely limit the increase of passengers travel by railways. 180 160 140 120 100 Gpkm 80 60 40 20 0 2013 2015 2020 2025 2030 2035 2040 2045 2050 Private Public   Fig. 12. Road passenger transport travel demand in the baseline scenario. Unlike passenger travel demand, freight travel demand is evenly distributed among different transportation modes. In the freight transport, road remains the main transport mode during the projected period. For example, the share of road freight in total freight travel demand ranges from 57% in 2013 to 62% in 2030 and then decline slightly to 58% in 2050 (Figure 13). Road 29    freight is followed by railways, inland waterways and pipeline during the projected period. These figures reflect, among others, the relative importance of passenger and freight transport in the country. While passenger travel patterns are more closely related to personal wealth and lifestyle changes, freight transport activities are closely connected to overall economic activity. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2013 2015 2020 2025 2030 2035 2040 2045 2050 Road Rail Inland waterways Pipeline Fig. 13. Share of freight transport demand by transport model in the baseline scenario. In recent years, both air passenger volumes, measured in number of passengers, and air freight volumes, measured in tons, grew steadily, closely connected to the country’s economic growth rate. It is projected that the trend will continue over the projected period. For example, air passenger volumes is projected to grow at an average rate of 1.1% per year during 2013-2050 to 17 million passengers in 2050, while air freight volumes is projected to grow at 1.4% per year over the same period to 55,000 tons in 2050 (Table 9). Maritime follows a similar trend as that of air transport during the projected period. 30    6. Concluding Remarks Using an end-use accounting model, this study projects sector-wise energy demand by end- uses and travel demand by transportation modes over the next 35 years. Although there is considerable variation in demand for energy between sectors, sub-sectors and end-uses, the country’s overall FED is projected to increase from 2013 (base year) to 2050. The outlook for FED differs to some degree across the three scenarios. In the baseline scenario, total FED grows at an average annual rate of 0.8% through 2050, rising from 21.6 Mtoe in 2013 to about 28.8 Mtoe in 2050. Demand grows slightly faster in the high scenario, by 1.2% per year, but slows to 0.5% per year in the low scenario, owing to in part to slower economic growth. As a result of the growing economy and rising income, the market for electric services and the stock of electronic devices and electric appliances are expected to continue to grow in the country. This will lead to faster projected rates of growth in demand for electricity than the total final energy in all scenarios. In the baseline scenario, electricity demand grows by 1.3% per year and it meets 20% of final energy needs in 2050, up from 16% in 2013. However, the difference in per capita electricity consumption in Romania and EU member countries remains very large. In 2040, per capita electricity consumption in Romania amounts to 3,239 kWh, while it is 5,710 kWh in Germany and 6,294 kWh on average in EU-28. At the sector level, the residential sector is presently the largest energy consuming sector in the country, accounting for a little more than one-third of total FED, followed by the industry, transport and services sectors. Interestingly, the share of Romanian residential energy demand in total FED in 2013 is the highest among all EU-28 countries. However, growth in energy demand across the sectors varies widely with time. The growth in residential energy demand is projected to slow with time, on average 0.1% per year to reach 7,911 ktoe in 2050 under the baseline 31    scenario, mainly due in part to declining population and using more efficient household appliances. Despite slow growth, residential energy demand accounts for more than one-fourth of total FED in 2050. At the end-use level, most of residential energy demand comes from space and water heating over the projected period, accounting for roughly three-fourth of total residential FED. The remaining one-fourth comes mainly from the use of electric appliances, cooking and lighting end-uses. Space heating, therefore, represents the largest opportunity to reduce residential energy demand by increasing energy efficiency of the heating equipment and by changing the energy-mix. Both in absolute and percentage terms, industry is projected to be the largest final energy consuming sector, surpassing the residential sector from 2025 onwards in all three scenarios. Industrial energy demand is projected to grow more rapidly at 1.2% per year, reaching 10,536 ktoe in 2050 in the baseline scenario. At the end-use level, thermal (heat) energy use is projected to account for roughly two-thirds of all industrial FED, mainly due in part to strong growth in energy-intensive manufacturing industries. The remaining one-third comes from use of electric motor and electric appliances, and for motive and thermo-electric requirements. Within industry, manufacturing is by far the largest energy end-user, accounting for 86% of total industrial FED in 2050 in the baseline scenario. In particular, manufacturing industries such as iron and steel, chemical and petro-chemicals, and non-metallic minerals, dominate total thermal energy demand. Although the share of services in total FED is relatively small, services FED is projected to grow at much higher rate of 1.7% per year, doubling in absolute value in 2050 from the 2013 level. Over the projection period, almost half of total services energy demand comes mainly from space heating in office buildings, hotels, and wholesale and retail stores. Besides 32    using improved and efficient heating equipment, there is also great potential in reducing energy demand by renovating office buildings. Currently, about one-fourth of total FED comes from transport sector. Energy demand in this sector steadily increases over 2013-2050, growing by 0.8% per year. In Romania, energy use in transport is heavily concentrated on road transport. Without any further policy measures, this trend is expected to remain the same over 2013-2050 period. In 2050, almost 90% of total transport energy demand comes from road transport in the baseline scenario. As for the travel demand in the country, road and railways are the main transport mode for passengers, while road, railway and inland waterways are the main transport modes for freight. Excluding air and maritime transport, the total passenger travel demand grows on average by 1.4% per year, reaching 167 Gpkm in 2050, while the total freight travel demand grows on average by 1.3%, reaching 94 Gtkm in 2050. Road transport dominates both passenger and freight projected travel demand during 2013-250. In particular, private passenger dominates road passenger travel demand, mainly due to rising incomes and increasing levels of private passenger vehicle ownership. Further, the lack of railway infrastructure coupled with declining population and growing number of vehicles during 2013-2050 period will most likely limit the increase of passengers travel by railways. Although reducing energy consumption in the transport sector is a particularly difficult challenge in the country, the Romanian government is encouraged to intensify its efforts to shift from private to public road and rail transport by promoting mass transit systems. 33    References Atanasiu, B., Economidou, M., Maio, J., 2012. The Challenges, Dynamics and Activities in the Building Sector and its Energy Demand in Romania. D2.1 of WP2 from Entranze Project. Available online at http://www.entranze.eu/files/downloads/D2_1/D2_1_Short_country_ summary_report_-final-Romania_july.pdf (accessed 01.06.2016). Bertoldi, P., Hirl, B., Labanca, N., 2012. Energy Efficiency Status Report 2012: Electricity Consumption and Efficiency Trends in the EU-27. JRC Scientific and Policy Report EUR 25405 EN. Available online at https://setis.ec.europa.eu/sites/default/files/reports/energy- efficiency-status-report-2012.pdf (accessed 01.06.2016). Bhattacharyya, S.C., Timilsina, G.R., 2010. Modelling energy demand of developing countries: Are the specific features adequately captured? Energy Policy 38, 1979-1990. Bianco, V., Manca, O., Nardini, S., Minea, A.A., 2010. Analysis and forecasting of nonresidential electricity consumption in Romania. Applied Energy 87, 3584-3590. CNP, 2015. Energy Balance Forecast (June 2015). The National Commission for Prognosis (CNP), Bucharest (in Romanian language). EC, 2008. NACE Rev.2: Statistical Classification of Economic Activities in the European Community. Eurostat Methodologies and Working papers. European Community (EC), Luxembourg. EU, 2015. Energy Balance Sheets: 2013 data. The European Union (EU), Luxembourg. EU, 2014a. Assessment of the 2014 National Reform Programme and Convergence Programme for Romania: Council Recommendation. SWD 424 Final. EU, Brussels. _______b. EU Energy, Transport and GHG Emissions Trends to 2050: Reference Scenario 2013. EU, Luxembourg. 34    Eurostat, 2015a. GDP and Main Components (nama_10_gdp), and Main GDP Aggregates per Capita (nama_10_pc). Available online at http://ec.europa.eu/eurostat/data/database (accessed 10.01.2015). ____________b. Energy Statistics, EU28: Energy Datasheets. Available online at https://ec.europa.eu/energy/sites/ener/files/documents/CountryDatasheets_June2015.pdf (accessed 10.01.2015). ____________c. EU Transport in Figures: Statistical Pocketbook 2015, Section 2.6. Available online at http://ec.europa.eu/transport/facts-fundings/statistics/pocketbook-2015_en.htm (accessed 10.01.2015). ____________d. Mean Consumption Expenditure (nama_co3_c). Available online at http://ec.europa.eu/eurostat/web/products-datasets/-/nama_co3_c (accessed 10.01.2015) ICEMENERG and ANRE, 2012. Energy Efficiency Policies and Measures in Romania in 2012: Monitoring of EU and National Efficiency Targets. Bucharest: Energy Research and Modernizing Institute (ICEMENERG) and Romanian Energy Regulatory Authority (ANRE). IEA, 2013. Energy Polices of EIA Countries: Germany, 2013 Review. International Energy Agency (IEA), Paris. IEA, 2014. World Energy Outlook: 2014. IEA, Paris. INS, 2015a. Monthly Averages of Total Expenditure per Household by Expenditure Categories and Main Social Categories. Online time series, National Institute of Statistics (INS), Bucharest. Available online at http://statistici.insse.ro/shop/?lang=en (accessed 11.10.2015). ________b. Number of Durable Goods per 100 Households. ________c. Energy - Balance of Electric Energy by Component Elements. ________d. Economic Statistics: National Accounts. 35    ITU, 2014. Measuring the Information Society Report 2014. International Telecommunication Union (ITU), Geneva. MECC, 2013. Romania’s Sixth National Communication on Climate Change and First Biennial Report. Ministry of Environment and Climate Change (MECC). Bucharest. Pardo, N., Vatopoulos, K., Krook-Riekola, A., Moya, J.A., Perez, A., 2012. Heating and Cooling Demand and Market Perspective. JRC Scientific and Policy Report EUR 25381 EN. Available online at https://setis.ec.europa.eu/sites/default/files/reports/Heat-and-cooling- demand-and-market-perspective.pdf (accessed 11.10.2015). World Bank, 2014. Summary of Sector Rapid Assessments and Recommendations for Incorporating Climate Actions in the 2014-2020 Sectoral Operational Programs in Romania. Component B Synthesis Report. The World Bank, Washington, DC. World Bank, 2015. World Development Indicators (Online Data Bank). The World Bank, Washington, DC. 36