The World in 2050: Four Forces Shaping Civilization's Northern Future

power, even renewable energy sources demand a lot of water.

Power plants—regardless of whether they run on coal, natural gas, uranium, biomass, garbage, or whatever—use water in at least two important ways: to make steam to turn a turbine and thus generate electricity; and to get rid of excess heat. The single greatest demand for water in the energy sector today is for the cooling of power plants. Over half of all water withdrawals in the United States alone, slightly more than for irrigating crops, are used for this purpose. That’s a half-billion acre-feet of water per year (enough to flood the entire country ankle-deep in water) to cool off our power plants. In some parts of Europe the percentage of water withdrawn for energy production is even higher.²³⁷

The total amount of water needed depends greatly on the fuel used, on plant design, whether the water is recycled, the type of cooling apparatus, and so on. But in all cases the volume of water needed to operate the power plant is large, even greater than the volume of fuel. This is why plants are sited next to water bodies or perched over large aquifers. It’s not uncommon to find a coal-fired power plant on a riverbank hundreds of miles from the nearest coal mine: It is cheaper to carry the coal to the water, rather than the other way around. The Three Mile Island nuclear power plant, site of the 1979 accident described in the previous chapter, really is on an island, stuck out in the middle of the Susquehanna River.

Power plants bite into water supply by reducing both its quality and its quantity. Water recycled back into a river is hotter than the water withdrawn, sometimes by as much 25°C.²³⁸ For plants located on large bodies of water like the ocean, this doesn’t introduce significant environmental harm. Putting hot water into a river or lake, however, degrades aquatic ecosystems for many reasons. Warm water holds less dissolved oxygen, slows the swimming speed of fish, and interferes with their reproduction. Desirable cool-water species like trout and smallmouth bass are replaced by warm-water species like carp.

The second problem is water consumption, meaning irrevocable water loss. Most power plants use “wet” cooling towers—or even open ponds—to deliberately evaporate water into the atmosphere, providing cooling in the same way that evaporating sweat cools your skin. Evaporation losses from power plants are much smaller than the total withdrawal but are still significant in water-stressed areas. In very dry places, it becomes increasingly difficult to guarantee enough water for cooling purposes at all.

In the first study of its kind, Martin Pasqualetti, a professor in the School of Geographical Sciences and Urban Planning at Arizona State University,²³⁹ scrutinized how much water consumption (i.e., evaporation) Arizona’s different energy technologies require in order to produce one megawatt-hour of electricity. What he found may surprise you:

Water Losses Embedded in Arizona Electricity Generation

From Pasqualetti’s data we learn that the water consumption of energy production is not only large, but varies tremendously depending on the type of energy being used. For example, a nuclear power plant evaporates about 785 gallons of water to generate one megawatt-hour of electricity, whereas natural gas power plants evaporate considerably less (especially modern combined-cycle plants, which evaporate about 195 gallons per megawatt-hour). This means that an average house in Phoenix, using twenty megawatt-hours per year, will unknowingly evaporate nearly 16,000 gallons of water if its electricity comes from a nuclear power plant, but only about 3,900 gallons if it comes from a combined-cycle natural gas plant. More virtual water.

To put that number into perspective, 15,000 gallons is roughly what a typical Phoenix household with irrigated landscaping uses in two weeks. So this “embedded” water is not an enormous amount, but still significant in such a dry place. But the big surprise here is that in terms of electricity generation, hydropower, of all things, is the worst water waster,²⁴⁰ followed by concentrated solar thermal (CSP) technology, then nuclear. Arizona does not grow biofuel crops, but other studies show biofuels are even worse than hydro in terms of water consumption.²⁴¹ Thus biofuels, hydropower, and nuclear energy, while hailed for being carbon- neutral (or nearly so), are worse even than coal when it comes to water consumption. Of the renewables, only wind and solar photovoltaics are truly benign—something, Pasqualetti points out, that would make solar photovoltaics more cost-competitive if the price of the saved water was taken into account.

The water-energy nexus works both ways. Examined in the opposite direction, energy is needed at every step along the way to deliver clean water to a house. Take again, for example, our typical Phoenix home, which consumes about an acre-foot of water per year. It requires two megawatt-hours of electricity—roughly 10% of the home’s total energy use—to pump that acre-foot uphill from the Colorado River some two hundred miles away, purify it, and pressurize it locally. But those megawatt-hours never appear on any electric bill; they are embedded within the water bill itself. Remarkably, almost all the cost of providing drinking water to Phoenix households is for the energy embodied within it, not for the water.

“Indeed,” says Pasqualetti, “water and energy are married to one another. Water is needed in electrical generating stations if they are to run efficiently. Energy, on the other hand, is needed to provide our houses with safe drinking water. How much of each commodity is needed to provide the other is something not well appreciated by the public.”²⁴²

It is something not well appreciated by politicians and planners either. Instead of recognizing this marriage between energy and water, their respective planning and regulatory agencies are almost always totally separate entities. “Energy analysts have typically ignored the water requirements of their proposed measures to meet stated energy security goals. Water analysts have typically ignored the energy requirements to meet stated water goals,” concluded a recent Oak Ridge National Laboratory report.²⁴³ Historically we have gotten away with this thanks to cheap water, cheap energy, or both. That cushion will continue to narrow as supplies of both tighten out to 2050.

One of the most widely anticipated outcomes of climate change is that the Hadley Cell circulation will weaken slightly and expand. This appears not only in a broad range of climate model projections for the future, but also from historical data extending three decades into the past.²⁴⁴ The effect of this is the spawning of more clouds and rain in the tropics, but even drier conditions and a poleward expansion of the two desert blast zones straddling both hips of the equator. Precipitation futures are notoriously difficult to project, but this is one of those things about which all the climate models agree. Put simply, many of the world’s wet places will become even wetter, and its dry places even drier.

Rainfall will increase around the equator, but decrease across the Mediterranean, Middle East, southwestern North America, and other dry zones. Rivers will run fuller in some places and lower in others. One highly regarded assessment tells us to get ready for 10%-40% runoff increases in eastern equatorial Africa, South America’s La Plata Basin, and high-latitude North America and Eurasia, but 10%-30% runoff declines in southern Africa, southern Europe, the Middle East, and western North America by the year 2050.²⁴⁵ Through the language of statistics, these models are telling us to brace for more floods and droughts like the ones in Iowa and California.

The Great Twenty-first-Century Drought?

Part of the explanation for the many floods and droughts that happened around the world in 2008 was that it was a La Nina year, meaning that sea surface temperatures (SSTs) in the eastern half of the tropical Pacific Ocean cooled off. This triggered, among other things, dry conditions over California, contributing to its ongoing drought (her counterpart, El Nino, is associated with warm SSTs and wetter conditions there). Through connections between the sloshing ocean and the atmosphere, this “Little Girl” had impacts on human water supply that reverberated worldwide.

My UCLA colleague Glen MacDonald, an expert in the study of prehistoric climate change, is deeply concerned that something like the 2008 La Nina could happen again—but persisting for decades rather than months. In fact, MacDonald and his students believe the American Southwest, in particular, could be struck by a drought worse than anything ever seen in modern times. From shrunken tree-rings and other prehistoric natural archives, they have assembled a growing body of evidence that the region suffered at least two extended “Perfect Droughts” (coined by MacDonald to describe periods when Southern California, northern California, and the upper Colorado River Basin all experienced drought simultaneously)