Since 1750, mankind has pumped some 150 billion tons of carbon dioxide into the atmosphere. Almost half that amount has been emitted since 2000, 9.9 billion tons of it in 2016 alone. U.S. President Donald Trump claims to think that links between carbon emissions and climate change are "a hoax," but if so, the hoax has taken in almost every scientist in the world. By burning fossil fuels, they believe, we have altered the chemistry of the air and the oceans. Last year was the first on record in which the atmosphere's carbon content never dropped below 400 parts per million, a level not seen for 800,000 years. The climate is warming and becoming more volatile, the ice caps are melting, and mass extinctions are underway. Another species of plant or animal disappears forever every 20 minutes or so.
Although this is shocking, it is hardly news, and strategic forecasters have been arguing for years over what its geopolitical consequences will be. But now it seems that some answers to this question are beginning to emerge.
Not If, but When
Most analysts think that the world's demand for energy will keep growing in the near future. But they also believe that as time passes, renewable sources of energy — hydroelectric, biomass and perhaps nuclear energy, but above all wind and solar — will replace fossil fuels, reducing carbon emissions. The main disagreements are over how quickly that will happen.
Most experts have concluded that change will come slowly. Analysts at Shell predict it will take a quarter of a century to reach a tipping point where the annual output of renewable energy matches the overall growth in demand, and the amount of carbon being released into the atmosphere levels off. Their peers at BP think 30 years will pass before then; those at Exxon say 75. The International Energy Agency (IEA) largely agrees, though it recently cut its estimate from 60 years to 35.
When compared with actual data, however, forecasters' records have consistently erred on the side of conservatism. Shell and BP base their estimates out to 2040 on the assumption that total energy demand will grow by 1.4 percent each year, even though the recent rate has been more like 1.0 percent. They, Exxon and the IEA assume that solar and wind energy supplies will grow between 5 percent and 9.5 percent each year, even though the recent rate has been above 15 percent, and that other non-fossil-fuel sources (nuclear, hydroelectric and biomass power) will expand between 1.4 percent and 1.9 percent annually, despite their recent performance of around 2.3 percent. All conclude that the demand for fossil fuels will continue growing through the 2020s and 2030s at 0.7-1.2 percent per year, though the recent trend has been 0.5 percent.
The IEA, at least, has been ready to admit its mistakes, even if it has been slow to correct them. Back in 2002, it predicted that in 2015 wind and solar sources would produce about 40 and 10 gigawatts (1 GW is equivalent to 1 billion watts) respectively in 2015. In 2005, it revised its estimates to 170 and 20 GW, hiking them up again in 2010 to 340 and 75. But the actual outputs were 430 GW of wind power and 240 GW of solar power. By that point, wind and solar energy accounted for one-third of the total global increase in energy demand.
Experts argue over why one another's guesses have been so wrong (and so consistently wrong in the same direction), but their mistakes remind us of an obvious point: No one knows what will happen. The best we can do is to make our assumptions explicit, map out their consequences and ask whether they are plausible. In this spirit, Kingsmill Bond, a new energy strategist for the investment research company TS Lombard, recently published the following chart showing when the tipping point described above will arrive under different assumptions about growth in demand and solar and wind power:
Which columns and rows we select depends on our guesses about future policies and attitudes, engineers' abilities to solve daunting technical problems, enthusiasm for investment in infrastructure and, of course, the performance of the global economy. But despite this heaping up of "ifs," "buts" and "maybes," neither the top and bottom rows nor the left- and right-hand columns seem very likely. That leaves us in the middle of the grid, with a tipping point coming in the 2020s or 2030s. Bond himself leans earlier rather than later, expecting wind and solar energy to grow at 20 percent and overall demand at 1 percent, in large part because of gains in efficiency. If he's right, the lines will cross in 2020, and by the early 2040s half of the world's energy will come from renewables. If so, then Rabah Arezki, the head of commodities at the International Monetary Fund, must also be right when he says we stand "at the onset of the biggest disruption in oil markets ever."
Consumers Keep Pace With Supply
But this is not the first time energy markets have faced massive disruptions, and we can learn some important lessons from the past. In 1865, just as the world's first fossil-fuel economy was hitting its stride, the economist William Stanley Jevons published a small book called The Coal Question, asking how Britain should manage its energy supply.
One of Jevons' biggest contributions was to think about fuel in terms of the energy return on energy input (ERoEI), which, he observed, had been rising for a century. Steam engines had been around since the 1690s but only became really important after 1776, when James Watt and Matthew Boulton separated the heating and cooling chambers and reduced coal consumption by 75 percent. In 1804, Richard Trevithick showed that lightweight, high-pressure engines were even more efficient, and from then on, the ERoEI constantly rose until, in the 1860s, a given lump of coal could do 10 times as much work as it had done in the 1760s. And yet, he observed, rather than falling by 90 percent, coal consumption had actually grown tenfold.
Jevons concluded that "It is a confusion of ideas to suppose that the economical use of fuel is equivalent to diminished consumption. The very contrary is the truth." The Jevons Paradox, as it's usually called, has applied throughout history: When energy is cheaper, people use more of it. Some environmentalists, such as the bestselling author Bill McKibben, envision a sustainable future in which humanity recognizes limits, learning to live more simply and consume less. But the Jevons Paradox suggests that this will not happen. Renewables will displace fossil fuels only if they drive the ERoEI down below the price of oil, natural gas and coal. There will be massive upfront costs for research and development and infrastructure, but once they have been paid, we will enter a new age of energy abundance.
Much of this bonanza will be consumed in forms that are already familiar. More than a billion people still lack reliable electricity and clean water, and providing these — surely not much for the world's poorest people to demand — will drive up global energy consumption by 15 percent. Even so, the biggest new markets will be in the developed world.
Back in 1980, a sixth of the energy consumed by the typical American household went into appliances and electronics. In the 2010s that share has more than doubled, despite huge improvements in the energy efficiency of most appliances. (Refrigerators use 60 percent less energy now than in 1980, while clothes washers use 70 percent less and lightbulbs use 80 percent less.) The explanation for this example of the Jevons Paradox is, of course, that we now have so many more electronic gadgets and use them so much. The average American spends close to two hours per day on a smartphone, and even though these devices are phenomenally efficient — an iPhone 5 consumes at most 41 cents' worth of electricity per year — there are now a billion of them, supported by a global network of 3 million data centers that consume about 1.5 percent of the planet's energy.
It seems fairly safe to assume that the expansion of electronics and particularly computing will continue — so long as renewables allow us to expand our energy use without poisoning the planet. At the opposite end of the energy chain from the pocket-sized smartphone, supercomputers will make particularly prodigious demands on global energy supplies. The Tianhe-2 machine in Guangzhou, which reigned as the world's fastest computer from June 2013 through November 2015, can perform 33.86 petaflops (or 33,860,000,000,000,000 operations per second) and would in fact manage 54.9 if it were set up in a more suitable space. To do this it sucks up 17.6 megawatts (1 MW is equivalent to 1 million watts) of energy itself, plus another 6.5 MW for its external cooling system. (Siting it in hot, humid Guangzhou perhaps was not such a good idea.)
The Jevons Paradox applies as much to supercomputers as it does to steam engines. The Sunway TaihuLight in China's Jiangsu province, which replaced the Tianhe-2 as the world's fastest machine in 2016, delivers 93 petaflops for 15 MW, while the more powerful IBM Summit — scheduled to come online in 2018 — should provide 150-300 petaflops for 10 MW.
The energy requirements of these supercomputers are huge but not out of reach. Even the Summit's 10 MW are less than one-millionth of the 12.3 terawatts the world generated last year — less, in fact, than the London-Paris Eurostar train. But some scientists suggest that we are still living in the infancy of computing and that it will only truly transform the world when machines can work at the exaflop scale (100 times as powerful as the IBM Summit). Such exaflop computers will have the power to think like a human brain, while a machine operating on the yottaflop scale (another million times faster than the exaflop) could, some scientists say, create the equivalent of a superorganism linking together the brains of everyone on Earth — though it might suck up the entire planet's energy supply, too.
We are now deep into science fiction territory. But even if none of these more extreme predictions come to pass, we can be pretty certain that the Jevons Paradox guarantees that if we do revolutionize energy capture in the 21st century, we will find ways to use everything we generate. And if that happens, we should expect it to transform humanity's place on Earth. A century or two from now, engineers (or, more likely, the computers they have created) will be desperately seeking ways to increase their energy budget beyond the meager amount emitted by the sun.
The Sun Rises in the East
Finally, a word on the geopolitics of a post-fossil-fuel world. Each system of energy capture has produced its own distinctive map of wealth and power. In the age of hunter-gatherers, when wild plants and animals supplied mankind's energy, the richest places on Earth were mostly maritime environments such as the Baltic Sea, southern Japan and the Pacific Northwest, where people could capture enough energy to live in year-round villages (though not enough to project power over more than a few miles).
The invention of agriculture, beginning in the Middle East 12,000 years ago, enabled people to raise energy capture by domesticating plants and animals. The richest places on Earth shifted to what I like to call the "Lucky Latitudes," stretching from China to the Mediterranean in the Old World and Peru to Mexico in the New. Here people could capture enough energy from their crops and herds to create cities of up to a million residents and project power across hundreds of miles, forming great empires.
The Industrial Revolution, beginning in Britain just 250 years ago, gave humanity access to the energy trapped in fossil fuels, and the seats of power shifted once again. Britain was not the only place with rich coal reserves, but by being the first to make use of them it was able to project its power globally in the 19th century, building an empire and commercial network the likes of which had never been seen before. Similarly, although the United States was certainly not the only place with a wealth of oil reserves, its early exploitation of this energy source gave it even greater global reach in the 20th century.
But what of the 21st-century shift from fossil fuels to renewable energy? The sun shines and the wind blows everywhere, and plenty of places are well suited to capture them. But it increasingly looks like wealth and power will shift from West to East. China has already passed the tipping point where renewables are coming online faster than demand is growing, and in 2015 consumption of fossil fuels fell by 1.4 percent. In the same year China accounted for 28 percent of global electric vehicle sales, 32 percent of solar panel installations and 47 percent of wind installations; in 2016 it overtook Europe in wind power and by 2020 will probably surpass it in solar power, too.
None of this is written in stone, of course, just as it wasn't inevitable that the United States would overtake Western Europe in the 20th century. But while America's dominance of the age of oil was not preordained, it was always highly likely. And now, a century on, we may already be able to see the outlines of geopolitics in a post-fossil-fuel world laid out before us.