Projecting World Energy Production
This page should contain a chart that is generated live from specified assumptions about the history and future of world energy production. If you see a graph near this paragraph, then all is well.
You might, on the other hand, see a banner for Java. If it does not go away after a minute, then your system does not have an up-to-date copy of Java installed. Instructions to install Java may be found at the bottom of this page.
Across the top of the chart are a series of tabs, labelled Coal, Oil, Gas, Wind, Solar, etc. Click any of these tabs to see the corresponding chart.
Below each graph you will see one or two adjustable sliders. Using your mouse, you can adjust the value of the parameter. The graph will change to reflect the new parameter values as you adjust the slider.
When there are more than two parameters that can affect the graph, then you will see a popup menu next to the slider. Use this menu to select which parameter you wish to experiment with. The menu has a scroll bar on its right hand side, so that you can access all 14 parameters.
The horizontal axis of the graph represents time, from 1850 to 2100. To see the graph unfold over time, first click in the checkbox labeled "Animate," and then click the "Start" button. The "Reset" button restores all parameters to their initial values, and blanks the graph. The "Help" button merely refers back to this page it exists only to help those who encounter the Java applet in third-party web pages.
For consistency and ease of comparison, all energy statistics are expressed in "Gboe" units. One "Gboe" of energy is the equivalent of a billion barrels of oil (Gboe = "Gigabarrels of oil equivalent"). One billion barrels of oil is equivalent to about 5.46 quads (i.e. 5.46 quadrillion BTU) of energy. For those accustomed to using other units of energy, one barrel of oil is approximately 1.26 million kilocalories, or 5.73 gigajoules, or 5.46 million BTU of energy. There are 7.33 barrels in one metric tonne of oil.
Fossil Fuels (coal, oil, natural gas)
Historical production data for coal, oil, and natural gas over the years 1965-2005 came from British Petroleum's Statistical Review of World Energy, 2006. Older historical data for the years 1850-1965 came from estimates provided by the History Database of the Environment, a publication of the government of The Netherlands.
The production rates of all three fossil fuels are projected into the future using Hubbert logistic curves whose parameters are chosen to match the 2005 rate of production and the 2005 cumulative (1850-2005) production. For each fossil fuel, there is a slider that represents the so-called "Ultimately Recoverable Resource" (URR) of the fossil fuel. The URR is the estimated total amount of the resource that can ever be obtained. The Hubbert curve of projected production depends crucially upon the URR parameter.
For this simulation I have relied upon the URR estimates of Jean Laherrère, as reported in his 2006 lecture notes entitled "Oil and Gas: What Future?" (available here). It is my opinion that Jean Laherrère produces the most reliable geological estimates to be found anywhere. His 2006 URR estimates are as follows. Coal: 4.4 Tboe (trillion barrels of oil equivalent); petroleum: 2.9 Tbo; natural gas: 2.2 Tboe. The figure for petroleum includes extra-heavy oil, tar sands, and bitumen.
I used Hubbert curves to project future fossil fuel production because I am persuaded that the long-term evolution of fossil fuel extraction follows trajectories in this general class, with short-term fluctuations caused by political influences and market price dynamics.
Biomass and Hydroelectric Energy
Biomass is the term for all sources of energy that come from recent biological activity: wood, bioethanol, biobutanol, biodiesel, and biogas. Although fossil fuels have their origin in plant biomass that was living hundreds of millions of years ago, they are excluded from the definition because they contain carbon that has long been out of the carbon cycle.
Historical consumption data for biomass energy came from estimates displayed in (Laherrère, 2006, Fig. 35). I have been unable to locate the ultimate source of Jean Laherrère's figures, so I cannot confirm that they are reliable. I have assumed that biomass as a source of energy is increasing at 5% per year, with a long-term maximum production rate of 15 Gboe per year.
Historical production data for hydroelectric energy (1965-2005) came from BP's Statistical Review of World Energy, 2006. I have assumed that hydroelectric generating capacity is increasing at 5% per year, with a long-term maximum production rate of 10 Gboe per year.
Although uranium constitutes an exhaustible resource, nuclear energy may in the future rely on the earth's abundant stores of thorium. For this reason I have assumed that the limiting factor to nuclear energy lies not in fuel supply, but instead in the public's willingness to tolerate the risks associated with nuclear power plants. I have made what I believe to be a conservative assumption that the growth rate of nuclear energy capacity is about 2% per year, and that the long-term maximum production rate is about 10 Gboe/yr.
Electric energy generated by windmills has only recently become economically profitable. Historical data from 1997 to 2006 came from the Wikipedia article on Wind Power. The world-wide growth rate for net installed wind generating capacity was 25% during the years 2005 and 2006. The World Wind Energy Association expects the compound growth rate between 2005 and 2010 to exceed 15% per year.
I have set up the simulation so that the growth rate declines from 25% in 2006 down to the setting indicated on the slider by 1% per year. For my initial estimate, I have assumed that wind generating capacity will slowly settle down to a growth rate of 8% per year, with a long-term maximum production rate of 15 Gboe per year.
Solar Energy (photovoltaics, CSPs, etc)
Although electricity generation by solar photovoltaic cells has been possible for decades, it is only now on the verge of true economic viability. At present the world-wide growth rate for net installed photovoltaic capacity is about 35% per year, with a lot of variability. I have set up the simulation so that the growth rate declines from 35% in 2005 down to the setting indicated on the slider by 1% per year.
Despite the high growth rate, the installed base of photovoltaic generating capacity is still very small, amounting to only 5.3 gigawatts as of the end of 2006, according to the Wikipedia article on Solar Power. I have assumed that photovoltaic and other means for generating electricity from the sun will slowly settle down to a growth rate of 8% per year, with a long-term maximum production rate of 15 Gboe per year.
Historical data for the world's population from 1950 to 2005 came from estimates made by the U.S. Census Bureau. For earlier years, i.e. 1850-1950, the Census Bureau provides a summary of scholarly estimates from a variety of independent sources.
Population projections came from the Population Division of the United Nations Department of Economic and Social Affairs. I used two of their three 2004 projections (medium and low). In the first, it is assumed that total fertility in the nations of the world will eventually stabilize at an average of 2.0 children per woman in her lifetime. This number is slightly less than the replacement rate (about 2.1), implying a long-term slow downward trend in population after the peak is reached in 2075.
The second projection, which I regard as much more likely, is based on the assumption that total fertility will eventually stabilize at 1.85 children per woman. This produces a population trajectory that peaks in 2040, and sharply declines thereafter. Both projections assume some modest improvements in the mortality rates for adults, yielding somewhat longer average lifetimes.
The slider labeled "Mean Children per Woman" (i.e. total fertility) in essence allows the user to pick any scenario between these two projections. Setting the total fertility slider to 1.85 produces the low projection, while a setting of 2.0 produces the medium projection. Any changes in this slider will be reflected in the Population curve, and also in the Per Capita chart.
As the future unfolds, I intend to keep this simulation up-to-date with the latest data. If better theories come along for predicting the depletion of fossil fuels, I will incorporate them as well.
Web browsers like Explorer and Safari use a plug-in known as the Java Runtime Environment (JRE) to allow them to display web pages that contain Java programming. The latest version of the JRE is 1.5, but I have written this interactive chart for version 1.3 so that it should run even on elderly (i.e. 5+ year old) computers.
1. Microsoft Windows
Prior to the appearance of Microsoft Vista in 2007, almost all Microsoft operating systems included the JRE. If you have Vista, or if this page cannot display the interactive graph for any reason, or if it just seems flaky and ill-behaved, then you can install the latest JRE by going to Sun's Java free download page. The JRE downloads in just a few minutes, and is installed with a couple of mouse clicks. Enjoy!
2. Macintosh and all varieties of Unix and Linux
From an aesthetic point of view, the interactive graph has a reasonably good appearance under Windows, Linux, and Unix. However, it really shines on the Macintosh. My congratulations to Apple for a superb job on their implementation of Java.
Copyright © 2007 by Loren Cobb. All rights reserved.