Energy [R]evolution 2010—a sustainable world energy outlook

One important underlying factor in energy scenario building is future population development. Population growth affects the size and composition of energy demand, directly and through its impact on economic growth and development. World Energy Outlook 2009 uses the United Nations Development Programme (UNPD 2009) projections for population development. For this study, the most recent population projections from UNDP up to 2050 in the medium variant are applied. Based on UNDP’s 2009 assessment, the world’s population is expected to grow by 0.86% per year on average over the period 2007 to 2050, from 6.7 billion people in 2007 to more than 9.1 billion by 2050. Population growth will slow over the projection period, from an average 1.2% per year during between 2007 and 2010 to 0.4% per year between 2040 and 2050. The population of the developing regions will continue to grow most rapidly. The Transition Economies will face a continuous decline, followed after a short while by the OECD Pacific countries. OECD Europe and OECD North America are expected to maintain their population, with a peak in around 2020/2030 and a slight decline afterwards. The share of the population living in today’s non-OECD countries will increase from the current 82% to 85% in 2050. China’s contribution to world population will drop from 20% today to 16% in 2050. Africa will remain the region with the highest growth rate, leading to a share of 22% of world population in 2050. Satisfying the energy needs of a growing population in the developing regions of the world in an environmentally friendly manner is a key challenge for achieving a global sustainable energy supply.

Economic productivity is a key driver for energy demand. Since 1971, each 1% increase in global GDP has been accompanied by a 0.6% increase in primary energy (Graus and Blomen 2008) consumption. The decoupling of energy demand and GDP growth is therefore a prerequisite for reducing demand in the future, if a continuing growth of GDP is to be achieved. Most global energy/economic/environmental models constructed in the past have relied on market exchange rates to place countries in a common currency for estimation and calibration. This approach has been the subject of considerable discussion in recent years, and the alternative of PPP (Nordhaus 2005) exchange rates has been proposed. Purchasing power parities compare the costs in different currencies of a fixed basket of traded and non-traded goods and services and yield a widely based measure of the standard of living. This is important in analysing the main drivers of energy demand or for comparing energy intensities among countries. Although PPP assessments are still relatively imprecise compared to statistics based on national income and product trade and national price indexes, they are considered to provide a better basis for global scenario development. In this study, we relied on PPP adjusted GDP estimates from the World Energy Outlook 2009 (IEA 2009a). However, as WEO 2009 only covers the time period up to 2030, the projections for 2030–2050 are based on our own estimates.

An increase in economic activity and a growing population does not necessarily have to result in an equivalent increase in energy demand. There is still a large potential for exploiting energy efficiency measures. The energy intensity of an economy in this study is defined as final energy use per unit of gross domestic product. Under the Reference scenario, we assume that energy intensity will be reduced by 1.25% on average per year, leading to a reduction in final energy demand per unit of GDP of about 15% between 2007 and 2020. This value compares well with the reduction of the energy intensity of EU-25 between 1990 and 2004 (see below). In comparison, the total energy consumption in the EU-25 grew at an annual rate of just over 0.8% over the period from 1990 to 2004, while GDP grew at an average annual rate of 2.1% during the same period (EEA 2010). As a result, total energy intensity in the EU-25 fell at an average rate of −1.2% per year (a total decrease of −16% between 1990 and 2004). Despite this relative decoupling, total energy consumption has increased by 12.0% overall in the period 1990–2004. Energy intensity declined over 1990–2000 (and continuously during 1996–2000) but has remained broadly stable since then. For the entire simulation period (2007 to 2050), an average annual decrease of the energy intensity of 1.25% results in a total reduction of 56% in these 53 years. Although the current energy intensity is very different from region to region, our study implicitly assumes that all regions will be able to reduce energy intensity to Japan’s level of 2007 within the next 30 years.

Under the advanced Energy [R]evolution scenario, it is assumed that active policy and technical support for energy efficiency measures will lead to an even higher reduction in energy intensity of almost 73% between 2007 and 2050. The advanced Energy [R]evolution scenario follows the same efficiency pathway, apart from in the transport sector, where a further reduction of 17% due to less vehicle use and lifestyle changes has been assumed. The increased share of electric vehicles in this scenario, with greater efficiency of electric drives, leads to a further decrease in final energy use. The energy intensity in an economy tends to decrease over time as a result of a number of factors, e.g.

The energy-intensity decrease in the reference scenario results from a mix of these three factors and differs per region and per sector. For the period 2005–2030, the energy-intensity decrease is taken from the WEO 2009. For the period 2030–2040 and the period 2040–2050, the development is based on the energy intensity per region and sector in the period 2020–2030 in WEO. However, we made a correction for the change in GDP growth rate per period to avoid a situation where the energy intensity decrease in the Reference scenario is larger than the economic growth rate. For the period 2030–2040 and 2040–2050, the energy intensity decrease is calculated by the following two formulae: where:

After defining energy intensities of the Reference scenario, technical potentials for energy efficiency improvement are estimated. In this step, a list is drawn up of energy savings options taken into account per sector. After that, the technical energy savings potential is estimated per measure. The technical potentials are based on:

The key assumptions for calculating technical potential are:

This study aims at calculating energy efficiency improvement by developing indicators for energy-intensity per sector and where possible by subsector.

The main energy consuming sectors are the industry and transport sectors, as well as “other sectors”, (residential sectors, services and agriculture; Graus and Blomen 2008) the subsector energy use and the selection of the measures per sector are discussed and shown in detail. Options are selected, which are expected to result in a substantial reduction of energy demand before 2050.

In the Reference scenario, total global energy demand is expected to increase from 305 EJ in 2007 to 352 EJ in 2050. The growth in the transport sector is projected to be the largest, with energy demand expected to grow from 82 EJ in 2007 to 158 EJ by 2050 (see Table 1). Demand from “other sectors” is expected to grow the least, from 124 EJ in 2007 to 198 EJ by 2050. Under the (basic) Energy [R]evolution scenario, however, growth in overall final energy demand can be limited to an increase of 12% up to 2050 in comparison to the 2007 level (341 EJ in 2050), whilst taking into account implementation constraints in terms of costs and other barriers. The increase of the energy demand in the transport sector is very small, while in the industry and other sectors, final energy demand increases by ca. 17% (resp. 15%) between 2007 and 2050.

Figure 1 shows the potential for energy efficiency measurements for the industry, transport and other sectors in 2050 for the different world regions, i.e. the difference between the energy demand in these sectors in 2050 in the Reference scenario and the respective demand in the basic E[R] scenario, normalised to the 2005 level of total energy demand in each world region. Furthermore, the remaining total energy demand in 2050 in the basic E[R] scenario (relative to 2005 levels) is shown.

The technical savings potential up to 2050 from all the measures described in (Graus and Blomen 2008) is summarised in Table 2. Since it was not clear what assumptions the IEA WEO Reference scenario was based on, they have assumed an efficiency improvement of 1% per year. Electricity use in the “other” sector was assumed to decline at the same rate as residential use (lighting, appliances, cold appliances, computers/servers and air conditioning). They have assumed a minimum energy efficiency improvement of 1.2% in the Technical scenario and 1.1% in their Revolution scenario, including autonomous improvements. For services and agriculture, they have assumed the same percentage savings potential as for the household sector all aggregated in “other sectors”. The new Reference scenario based on WEO 2009 data now includes a lower level of energy demand in the residential sector. Therefore the savings used in the new Energy [R]evolution scenarios are lower than the figures shown in Table 2. The resulting final energy demand reduction for the Energy [R]evolution scenarios compared to the Reference scenario is shown in Table 3 for each world region.

Bio energy is an important storable renewable energy source. However, the use of bio energy is controversial and a sustainable fuel supply chain is crucial. The limited availability of sustainable bio energy requires very efficient use especially for heating and cooling in cogeneration power plants, where overall efficiency is far higher than biomass use in the transport system (in combustion engines). As a response to the controversial discussion on the availability of biomass resources, a study on the global potential for sustainable biomass was commissioned as part of the Energy [R]evolution 2008 project (Seidenberger et al. 2008). The German Biomass Research Centre, the former Institute for Energy and Environment, compiled research into worldwide energy crop potentials in different scenarios till 2050. Additionally, scientific literature on the status quo of worldwide potential studies and the state of the art of remote sensing for investigation of biomass potentials was compiled by Seidenberger et al. (2008). As the results of the Seidenberger et al. (2008) study are not publicly available, the key results of this study are summarised in the following paragraphs:

Residues are products from forestry, agricultural waste and by-products from food production as well as waste from wood products and animals. Residues can be dry matter, e.g. wood chips as well as wet matter, e.g. animal waste. The share of each residue is a fraction of the total amount of residual biomass can vary in different regions and is mainly dependent on the population, living standards and the methods and intensity of the agricultural and forestry production in the particular region. Several studies analysing long-term residue potential in a more or less detailed way are available. A direct comparison of the studies is difficult, since the baseline assumptions are different.

Following Seidenberger et al. (2008), we used results from Dessus et al. (1993) for 2020, as it is the only study with region-specific residue potentials for 2020. For 2050, biomass residuals potential is based on Smeets et al. (2007) as the authors have defined sustainability criteria in their assessment. Moreover, Smeets et al. (2007) offer a relatively high level of transparency and traceability from the methodological point of view. Nevertheless, it must be pointed out that the calculated potentials seem to be conservative and were partly converted from the original aggregation necessary for our scenario analysis, which is listed below in Table 4. Because of the lack of data, the Asian region is most problematic.

Besides the utilisation of biomass from residues, the production of energy crops in agricultural production systems is of controversial. Therefore, the technical potentials of energy crops were calculated assuming that the demand for food takes priority. In a first step, different scenarios for the demand of arable land and grassland for food production were calculated for each of 133 countries.

The needs and surpluses of agricultural areas are balanced between the countries of the groups EU-27, other European countries, North America, Central America, South America, Oceania, Asia and Africa to estimate the area available for the cultivation of energy crops in each world region. In a next step, the surpluses of agricultural area in each world region are classified as arable land and grassland. On grassland hay and grass silage are produced, on arable land fodder silage³ and short rotation coppice⁴ (SRC) are cultivated. Silage of green fodder and grass are assumed for biogas production, wood from SRC and hay from grasslands are assumed for the production of heat, electricity and synthetic fuels (BtL⁵ or ethanol from lignocelluloses).⁶ Country specific yield developments are taken into consideration.

As a result, the global biomass potential from energy crops in 2050 was estimated to range from 6 EJ in the Sub 1 scenario to 97 EJ in the BAU scenario (see Fig. 2). In comparison to the BAU scenario, potentials decrease clearly in the Basic and Sub 1 scenario, and the lowest potentials exist in the Sub 1 scenario. The considerable higher demand of agricultural area in the Sub 1 scenario compared to the Basic and the BAU scenario is due to an ecological orientated agriculture with less fertilizer and less pesticide and therefore lower specific yields. In the Sub 2 scenario, considerable higher energy crop potentials can be released by changing the human food pattern, reducing meat consumption and consequently, the area necessary for fodder production. Also in the Sub 3 scenario, considerable potentials can be realized, in most cases even higher than in the BAU scenario. The most important country for the differences between the scenarios in 2050 is Brazil. In the BAU scenario, big agricultural areas are released by deforestation in Brazil, whereas in the Basic and Sub 1 scenario, this deforestation does not occur anymore. Consequently, no additional agricultural area for energy crops is available in Brazil in these two scenarios. In contrast high potentials are available in the Sub 2 scenario as a consequence of the reduced meat consumption of the Brazilian. Because of high population and low quantity of agricultural area, no area surpluses for energy crop production are available in Central America, Asia and Africa. However, the EU, North America and Australia have relatively stable potentials.

The Basic and Sub 3 scenario are of particular importance, since the Basic scenario would be the “minimum solution” for future agriculture. The Sub 3 scenario demonstrates the development of an ecological orientated agriculture. But such a development is only realistic, if the eating behaviour changes. Otherwise, the higher demand of food crops from an increasing world population cannot be compensated. The results of the calculation show that the availability of biomass resources is driven by different factors (as evident in the boundary conditions set in the different scenarios above), which do not only affect the global food situation but also the conservation of natural forests and other biospheres. So, the assessment of future biomass potentials is the starting point for the discussion about the integration of bio energy into a renewable energy system.

The total global biomass potential (energy crops and residues) in 2020 ranges from 66 EJ (Sub scenario 1) to 110 EJ (Sub scenario 2). For 2050, scenario results range from 94 EJ (Sub scenario 1) to 184 EJ (BAU scenario). Those numbers are conservative calculations and have an estimated uncertainty, especially for in 2050, of a factor of two. Reasons for this uncertainty are due to the unknown consequences of climate change on agricultural production, possible changes of the worldwide political and economical dynamics, a higher yield increase as a consequence of a change in agricultural techniques and/or the faster development in plant breeding. The global potential for biomass residues is estimated to be 88 EJ in 2050 (see Table 4). With a biomass consumption of 88.7 EJ in 2050, the Energy [R]evolution scenario complies with the most stringent requirements towards sustainable biomass use.

The recent dramatic fluctuations in global oil prices have resulted in slightly higher forward price projections for fossil fuels. Under the 2004 “high oil and gas price” scenario from the European Commission, for example, an oil price of just $34 per barrel was assumed in 2030. More recent projections of oil prices by 2030 in the World Energy Outlook (IEA 2009a) range from $₂₀₀₈ 80/bbl in the lower prices sensitivity case and up to $₂₀₀₈ 150/bbl in the higher prices sensitivity case. The Reference scenario in WEO 2009 assumes an oil price of $₂₀₀₈ 115/bbl. Since the first Energy [R]evolution study was published in 2007, however, the actual price of oil has moved over $100/bbl for the first time, and in July 2008 reached a record high of more than $140/bbl. Although oil prices fell back to $100/bbl in September 2008 and around $80/bbl in April 2010, the projections in the IEA reference scenario might still be considered too conservative. Taking into account the growing global demand for oil, we have assumed a price development path for fossil fuels based on the IEA WEO 2009 higher prices sensitivity case extrapolated forward to 2050 (see Table 5). As the supply of natural gas is limited by the availability of pipeline infrastructure, there is no world market price. In most regions of the world, the gas price is directly tied to the price of oil. As a consequence, gas prices used in this study are assumed to increase to $24–29/GJ by 2050. Additional price projections for biomass considered that biomass from energy crops are mainly available in the industrialised countries, especially in Europe as calculated by Seidenberger et al. (2008). Thus, biomass prices in Europe are assumed to be much higher than prices for residual biomass in the other regions.

Assuming that a CO₂ emission trading system will be established across all world regions in the longer term, the cost of CO₂ allowances needs to be included in the calculation of electricity generation costs. Projections of emissions costs are even more uncertain than energy prices, however, and available studies span a broad range of future estimates. As in the previous Energy [R]evolution study, we assume CO₂ costs of $10/tCO₂ in 2015, rising to $50/tCO₂ by 2050. Additional CO₂ costs are applied in Kyoto Protocol Non-Annex B (developing) countries only after 2020. The cost projections for CO₂ are relatively conservative due to the fact that a global emission trading system requires a strong and ambitious mandatory framework to reduce global energy related CO₂ emissions. However, the UNFCCC conference in Copenhagen in December 2009 failed to agree on such legally binding targets, and a global emission trading scheme will require several more years of negotiations.