Global energy technology transitions.

Past energy technology transitions provide lessons for current efforts and can inform future visions of energy technology change in the twenty-first century.

The global energy system is a planetary-scale complex network of energy converters and energy flows. Figure II.1 illustrates the associated global flows of “exergy”, that is, the energy available to be used, at the most aggregate level, from extraction at the primary level through the secondary, final and useful energy levels. It illustrates the dominance of fossil fuels and the low overall efficiency of the global system. It should also be noted that most greenhouse gas mitigation efforts focus on electricity supply, even though most losses are incurred from the final to the useful level. Useful energy is divided into motion (transport and machines), heat (mainly buildings) and non-energy (dominated by six materials). The underlying global reference energy system of interlinked energy technologies is even more complex. As the overall system is more than the sum of its components, for most purposes neither specific energy technologies (such as a wind power plant) nor specific parts of the system (such as renewable energy) should be analysed in isolation.

Over the past 200 years, global energy use has grown by a factor of 25-530 exajoules (EJ) in 2009, compared with a sevenfold increase in population, driven by demand for higherquality energy services made possible by underlying energy technology change. Throughout the twentieth century, total energy use in developed countries had been much higher than in developing countries. In recent decades, however, energy demand in China, India and other emerging economies has grown rapidly, so that by 2009, more than half of primary energy was used in developing countries. This share is expected to continue to increase to at least two thirds over the coming decades. At present, 2.7 billion people continue to rely on traditional, non-commercial fuels typical for pre-industrial societies. They use only between 15 and 50 GJ of primary energy per capita, delivering about 2-5 GJ of per capita in useful energy services. Growth of per capita energy use in developing countries has accelerated since 1975, whereas use in developed countries has stagnated (figure II.2).

 There are persisting differences between the development trajectories of countries, spanning the extremes of highly energy-intensive and highly energy-efficient. Initial differences in resource endowments or social configurations can be perpetuated over time by differences in economic activity, technology adoption rates, consumption patterns and infrastructure, which shape the direction of path-dependent energy technology change. At 1990 levels of energy conversion efficiency, a minimum level of 40-50 GJ primary energy per capita would be associated with a decent quality of life. Cross-country evidence suggests that, typically, no additional human development gains are obtained through primary energy use above 110 GJ per capita at prevailing conversion efficiencies. Improving overall global energy conversion efficiency from the present 11 per cent to 17 per cent would result in provision of the same level of energy services using 70 GJ of primary energy per capita.

Two major energy technology transitions have shaped the structure of the global energy system and the qualitative dimension of energy use since the onset of the industrial revolution (Nakicenovic, Grübler and McDonald, 1998). The first, associated with the emergence of steam power relying on coal (Landes, 1969), took more than a century to unfold (figure II.3). The second, characterized by the displacement of the coal-based steam technology cluster by electricity and petroleum-based technologies, is only about halfcomplete, with 2.7 billion people still lacking access to modern energy services (Global Energy Assessment, forthcoming). These transitions towards higher-quality energy fuels took place through successive substitutions going from traditional fuels to oil, gas and nuclear. At the global level, these substitutions had occurred at intervals of 70-100 years for the 250 years until 1975. Since 1975, however, this process has slowed to a global transition time of about 250 years, primarily as a result of government interventions and politically induced high and volatile oil prices. In addition, at the level of plants and units, there is no evidence of any differences in the speed of energy technology change across a wide range of technologies since the nineteenth century. 

The historical energy technology transitions have been characterized by a number of stylized patterns: 

• End-use applications drive supply-side transformations 

• Quality/performance dominates cost in the initial market niches

• Energy technologies do not change individually but in clusters, with “spillovers

• Time periods for energy technology change are decades, not years

• Experimentation and learning precede “upscaling” and widespread diffusion

• The size and the rate of expansion of energy conversion capacity are inversely related 

• Diffusion in late “adopter” regions is faster than in initial “innovator” regions, but maximum market penetration levels are lower 

• Both sufficient time and resources are needed for energy technology learning

From 1950 to 1990, energy-related global greenhouse gas emissions increased by about 3 per cent per year, mainly owing to increases in population (+1.8 per cent) and incomes (+1.9 per cent), the effect of which was moderated by lower energy intensity consumption patterns (-0.3 per cent) and better, lower-carbon technologies (-0.4 per cent) (Waggoner and Ausubel, 2002). Policy has typically focused on technology as the main lever for reducing emissions and it is indeed a powerful driver. Future energy technology change will be as important for determining future greenhouse gas emissions levels as long-term demographic and economic developments over the course of the twenty-first century (Roehrl and Riahi, 2000). Alternative technology strategies result in a divergence of emissions levels only gradually, after several decades or more, owing to the long lifetimes of power plants, refineries, buildings and energy infrastructure (Grübler, 2004); but nearterm technology and policy decisions will have sown the seeds of subsequent divergences, translating into different environmental outcomes as new technologies gradually replace older ones. Scenario analysis has helped to identify robust energy technology portfolios across a wide range of assumptions with respect to energy demand, resource constraints and availability and cost of technologies, and the extent of greenhouse gas constraints (Roehrl and Riahi, 2000; Riahi, Grübler and Nakicenovic, 2007; Grübler and Riahi, 2010). Figure II.4 illustrates the contributions of energy technologies to greenhouse gas emissions reductions, under a high-emissions baseline scenario, required for stabilization at a concentration of 550 ppmv CO2 e by 2100. The top two “mitigation wedges” show the contributions (in annual gigatons of carbon (GtC)) of (supply-side) carbon intensity improvements and (demand-side) energy intensity improvements in the baseline relative to a “frozen” state of technological development in 2000. This difference illustrates the innovation challenge of incremental energy efficiency improvements. Popular claims that “the technology exists to solve the climate problem” reflect the idea that no necessarily disruptive revolutionary changes (for instance, to nuclear fusion) would be needed, but such beliefs underestimate the challenge of achieving continued incremental improvements in line with historical trends. The ranking of these mitigation wedges is quite robust across scenarios, with energy conservation and efficiency accounting for more than half the emissions reductions. Indeed, across the wide range of scenarios, energy efficiency contributed about 59 per cent of the cumulative greenhouse gas emissions reductions from 2000 to 2100, compared with the much lower contributions of renewables (18 per cent), nuclear (9 per cent), fossil fuels (6 per cent) and other means (8 per cent) (Riahi, Grübler and Nakicenovic, 2007). Several global scenarios have explored feasible levels of lower per capita energy use and low greenhouse gas emissions that do not compromise economic development. For example, the GEA efficiency scenario shows that the world’s average primary energy use per capita can decrease from 71 to 63 GJ from 2010 to 2050, with average per capita gross domestic product (GDP) still tripling (in constant dollar terms). This implies an improvement in eco-efficiency by a factor of 3.2, which is almost as ambitious as the “factor 4” goal of doubling wealth and halving resource use, originally suggested by von Weizsäcker, Lovins and Lovins (1998). Thus, even world average levels of energy use of less than 70 GJ per capita would be achievable by mid-century, in line with the target suggested by the present Survey (see below).