But as for world demand for energy, we would likely expect that to increase in the present century. From 1900-2000 world energy use doubled on average every 25 years, a continual growth rate of 2.8%/year. In recent decades that growth rate slowed slightly, but still seems to be well over 2%/year, and the increase from 2002 to 2003 was 2.9%, according to the BP statistics.

Can we expect continued growth of 2 - 3% per year in the present century? What are the underlying causes of this growth in energy use, and why does the world need so much?

Vaclav Smil, in Chapter 3 of Energy at the Crossroads, "Against Forecasting", makes a strong case that all such projections are doomed, and complex models are likely not worth the effort, when simple estimates give about the same numbers. Circumstances change, technology changes the situation sometimes very rapidly, and there are always surprises that could throw off any forecast.

Nevertheless, the elements of the energy demand equation can be simply characterized: world population, world per capita gross domestic product (GDP), and world average energy intensity (energy use per unit of GDP). Each of these elements has their own rate of change with their own degree of unpredictability.

Population is perhaps the most predictable of the three elements of energy demand, or at least the most well studied. From the UN World Population Prospects data, projections of a continued decline in fertility and existing demographics show a range of about 7.3 to 8.3 billion people in 2025 (growth rates of 0.7 to 1.2%), and 7.5 to 10.5 billion people in the year 2050 (growth rates of 0.4 to 1.1% over that period). Barring cataclysmic disasters on a global scale, those ranges ought to be reasonably reliable.

Projections of economic growth clearly depend on the parameters of energy supply and regulatory and environmental constraints in a complex feedback. They also quirkily depend on the unit of measurement - how does one accurately measure "constant dollars" in a world with inflation and considerable year-over-year exchange rate variations, which may differ from purchasing-power parity measures?

Using the International Monetary Fund's figures, world per capita real ("constant prices") GDP rose through the 20th century at a rate of about 1.6%/year. Continued growth at this rate or higher could spread prosperity well beyond the developed nations, and seems like a good thing. Some recent years have seen world GDP per capita grow at rates of 3% or more.

The 1992 Intergovernmental Panel on Climate Change (IPCC) developed a scenario that is now referred to as the "Business as Usual" (BAU) model; this assumed a total GDP growth rate of 2.9% through 2025, and 2.3% after that point, so 1.7% and 1.2% in per capita growth respectively, relative to the high-end UN population growth numbers. The IPCC BAU scenario also assumed a steady 1% annual decline in energy intensity, the ratio of energy required per dollar of economic activity.

Is it justified to assume a continual decline in (primary) energy intensity? Averaged over the entire 20th century, energy intensity improved at a net rate of only about 0.2%/year; the number rose and then fell through the century. Energy intensity is often viewed as an approximate measure of energy efficiency - efficiency efforts were certainly responsible for much of the recent decline. However, economic shifts to less energy intensive activities also played a major role, since information and services tend to use a lot less energy than manufacturing. These shifts also make comparisons between nations more difficult. Vaclav Smil's "Energy at the Crossroads" devotes considerable space to discussing the energy intensity puzzle.

The International Institute for Applied Systems Analysis (IIASA) has published a number of further models, assuming energy intensity improvements at a rate of between 0.8 and 1.4% per year over the next 50 years; these and other different assumptions make a huge difference after 50 years of geometric improvement. Are such improvements actually achievable?

Even the slowest growth models predict energy use levels in 2050 of some 1.5 times as high as today; if growth is high and energy intensity doesn't improve much, energy use in 2050 could be over 3 times what it is now.

Putting this all in a table of simple scenarios:

	2000	2025	2050	2100
Population	6.1 billion	7.3-8.3 billion	7.5-10.5 billion	7-14 billion
World GDP (1990 US$/year)	$31 trillion	$50-70 trillion	$70-140 trillion	$120-500 trillion
World Energy use (quadrillion Btu/year)	400	450-850	450-1600	400-5000

Even under the most optimistic scenarios, with the best likely improvements in energy intensity and efficiency, demand for energy remains as high as it is now through the rest of this century with no let-up. In the worst case energy demand by 2100 skyrockets to over 10 times current demand. As Smil suggests, these projections are not prophecies - one circumstance or another could push the world outside even the wildest of these estimates. Nevertheless, the energy challenge is clear - and the opportunities are immense for any technology that can help meet it.

References:
Vaclav Smil, Energy at the Crossroads: Global Perspectives and Uncertainties, MIT Press (2003).
United Nations Population Division, World Population Prospects: The 2002 Revision Population Database - http://esa.un.org/unpp/
International Monetary Fund World Economic Outlook, http://www.imf.org/external/pubs/ft/weo/2000/01/
IIASA-ECS Models - Nakicenovic et al - see http://www.iiasa.ac.at/Research/ECS/docs/models.html
Innovative Energy Strategies for CO₂ Stabilization, edited by Robert G. Watts, Cambridge University Press (2002).

Comments

a bit of unit conversion Written by alizard on 2005-01-29 10:07:48 Most of us, I think are a bit more used to electrical energy measurements than BTU for overall energy consumption measurements.     Assuming that I didn't drop a decimal point:  1000 quadrillion BTUs/year=  1,000,000,000,000,000,000 Btu = 293,100,000,000,000 Kw-hrs = 293,100 GWh    (using the Length, Mass and Metrology conversions calculator at http://mdmetric.com/tech/lthmasscvt.htm)

almost Written by apsmith on 2005-01-30 01:37:24 Actually, that should be 293,100 TWh, not GWh. Otherwise righ - see the notes here on units for some more conversions. Basically, one quad is the same as the metric exajoule - 1 quad = 1.055 EJ or 293 TWh.    Various sources quote the numbers in all manner of forms - quadrillion Btu's seems to be favored by the US Dept. of Energy, others seem to like numbers in millions of tonnes of oil equivalent. For each different energy source you'll also see numbers in their own units - short tons of coal, trillions of cubic feet of natural gas, etc. It makes things very hard to compare.    The other problem is these are primary energy use numbers; for electricity there's a heat rate factor of about 3 that enters in, so 293,100 TWh is actually closer to 100,000 TWh(electric), because of the roughly 30% conversion efficiency from burned fuel energy to electric energy. If that makes sense - I wish there was a better way to present the numbers!

Only registered users can write comments.
Please login or register.

Powered by AkoComment 2.0!