Climate Change: The Science of Global Warming and our Energy Future

AS A WORLD SOCIETY, we have some control over our future climate by how we produce and consume energy simply because energy production accounts for about 80 percent of carbon dioxide (CO₂) emissions. Most of the rest comes from deforestation and land-use change related to agriculture, so it is in tackling energy use that we will make the biggest dent in carbon emissions.

A look at world energy consumption in 2005 by fuel type reveals what probably most of us already suspected. Fossil fuels—oil (37 percent), coal (27 percent), and natural gas (23 percent)—together accounted for 87 percent of the world’s energy (figure 10.1). The total consumption globally that year was 488 billion billion joules (or exajoules; for an explanation of the units and measures applied to energy, see table 10.1).

Carbon dioxide emissions from fossil-fuel burning for 2005 were about 29 billion metric tons (gigatons, see table 10.1). (Another 7 gigatons are thought to be generated by deforestation and agricultural activities, bringing total anthropogenic emissions to about 36 gigatons of CO₂ per year.) Recall from chapter 9 (and the estimates in table 9.1) what it would take to limit the increase in average global surface temperature to less than 3°C (5.4°F) above preindustrial temperature: cutting CO₂ emissions to 2000 levels or lower by 2050, corresponding to emissions of about 25 gigatons or less of CO₂ per year.¹ Furthermore, if energy consumption continues to grow at its present pace and energy production patterns do not significantly change, the emissions rate from burning fossil fuels will grow more than 50 percent to well more than 40 gigatons per year by 2030 (figure 10.2).² According to table 9.1, even if emissions were to stabilize but not decrease by then, global mean temperature would increase by more than 5°C (9°F) sometime later in the twenty-first century.

What are the solutions to this dilemma? An examination of CO₂ emissions by sector reveals that electricity generation accounts for about one-third of present day CO₂ emissions. More important, emissions from this sector are growing far more rapidly than emissions from other sectors (figure 10.3). (The emissions rate from transportation is also growing rapidly, but future transportation will likely derive more of its power from electricity and alternative fuels.) Therefore, although avoiding significant climate change will require mitigation efforts on other fronts as well, the crux of the problem is how we are going to meet the world’s electricity needs without bringing on catastrophic climate change.

TABLE 10.1 SOME COMMON UNITS OF ENERGY AND POWER
*Prefixes*

Kilo (k)	10³
Mega (M)	10⁶
Giga (G)	10⁹
Tera (T)	10¹²
Peta (P)	10¹⁵
Exa (E)	10¹⁸

*Units and Some Common Amounts*
Joule (J) = basic unit of energy exajoule = 10¹⁸ joules British thermal unit (Btu) = energy needed to heat 1 pound of water 1°F = 1,055 joules 1.055 exajoule = 10¹⁵ (1 quadrillion [quad]) Btu Toe = tons of oil equivalent = 41.868 × 10⁹ joules 1 million toe = 41.868 petajoule Watt (W) = unit of power (work) = energy per unit time = 1 joule/sec kilowatt = 1,000 watts, megawatt = 10⁶ watts, gigawatt = 10⁹ watts Watt hours (Wh) = energy = 1 W delivered over 1 hour = 1 joules/sec × 3,600 sec/hr = 3,600 joules 1 kilowatt hour = 3.6 × 10⁶ joules, 1 megawatt hour = 3.6 × 10⁹ joules, 1 gigawatt hour = 3.6 × 10¹² joules; 1 kilowatt hour = 3,413 Btu Metric ton (t) = 1,000 kilograms

Emissions are expressed as gigatons (10⁹ metric tons) of CO₂ per year. The business-as-usual projection is not a portrayal of the future. It simply suggests what might happen under the assumptions that no major conflicts or other events will disturb the world economy and that no changes in government policies will occur. (Data from Energy Information Agency, http://www.eia.doe.gov/)

Most electricity, as it turns out, comes from coal (40 percent), natural gas (20 percent), nuclear power (15 percent), and hydropower (16 percent), with oil (6.9 percent) and renewable sources other than hydropower (2.2 percent) playing only minor roles.³ Hydropower resources are limited and thus cannot accommodate the growth in demand. Renewable sources include geothermal, biomass (wood, landfill gas, agriculture by-products, and the like), wave and tidal, solar, and wind power. According to a recent report, even with major investment geothermal power is unlikely to account for more than 5 percent of electricity needs in the United States by 2050,⁴ and presumably that is true for the rest of the world as well; biomass, which currently accounts for about 20 percent of electricity generated in the United States from renewable sources (excluding hydropower), cannot substantially expand; and wave and tidal power are insignificant as a proportion of total world electricity capacity. This leaves wind and solar power, both of which are on the verge of major expansion. Central to the production of “clean” electricity, then, are (1) more efficient burning of coal and natural gas in combination with CO₂ capture and sequestration, (2) nuclear power, and (3) solar and wind power, so they are the focus of this final chapter.

Emissions are expressed as gigatons of CO₂ per year. The category deforestation includes burning of wood for fuel; other includes domestic surface transportation, nonenergy use of fuels, cement production, and venting/flaring from oil production; international transport includes aviation and marine transport. (After Rogner et al. 2007:fig. 1.2)

When it comes to coal, there are certain inescapable realities. Coal is cheap and abundant, especially in the countries that are and will continue in the foreseeable future to be the largest energy consumers (the United States, China, and to a lesser extent India). Furthermore, the production of energy from coal is supported by a large, extant infrastructure. Coal will therefore continue to provide a large proportion of the world’s energy in the decades to come. The consumption of coal has dramatically increased in the past several years. Indeed, in 2006 consumption jumped by an astronomical 4.5 percent, compared with the 10-year average growth rate of 2.8 percent before this year.⁵

The burning of coal presents significant problems. First, coal produces more CO₂ per unit energy than do other fossil fuels because it has a much higher ratio of carbon to hydrogen. In comparison with natural gas, for example, coal burning generates about 170 percent more CO₂ for an equivalent amount of energy, depending on coal type. Second, coal burning produces sulfur dioxide, nitrogen oxides (collectively, NOx gases), particulates, mercury, cadmium, uranium, and other toxic pollutants. Third, in addition to having negative environmental, health, and safety consequences, mining and transport of coal are themselves energy intensive.

“The scale of the enterprise is vast.” So states a recent study on energy from coal.⁶ To emphasize the point, the study begins with some startling statistics. Among them are that the equivalent of more than five hundred 500-megawatt (see table 10.1) coal-fired power plants in the United States are producing in aggregate about 1.5 gigatons of CO₂ per year, and, more striking, China has been constructing the equivalent of two 500-megawatt coal-fired power plants per week, a capacity comparable to the entire annual power grid of the United Kingdom.

World coal reserves (that is, the amount for which recovery is reasonably certain) are estimated to be 905 gigatons, an amount sufficient to power the world for the next 200 years.⁷ Just four countries—the United States (27 percent), Russia (17 percent), China (13 percent), and India (10 percent)—hold 67 percent of the recoverable reserves, with most of the remaining reserves concentrated in just a few other countries (table 10.2). This situation makes coal even more localized than oil in terms of geographical distribution, although it is concentrated in more politically stable parts of the world. Coal has another important advantage: it is inexpensive. In 2007, electricity generated from coal cost anywhere from about 10 to 30 percent of that generated by oil and natural gas.

Projections have coal consumption increasing rapidly over the next two decades, due almost entirely to growth in China and the United States (figure 10.4). In China, annual consumption is projected to rise from 43 to 100 exajoules worth of coal from 2004 to 2030, reflecting a combination of the country’s rapid economic growth and its lack of indigenous oil and natural-gas reserves. China’s coal-fired electricity-generating capacity in 2004 was about 271 gigawatts; by 2030, that capacity will have nearly tripled to 770 gigawatts, consuming 59 exajoules worth of coal annually. Much of the rest of China’s coal is used by the industrial sector—for example, in the production of steel and pig iron (China is the world’s leading producer of both). By 2030, China’s industrial coal use is expected to reach 39 exajoules, nearly 40 percent of the country’s total consumption.

TABLE 10.2 WORLD’S RECOVERABLE COAL RESERVES IN GIGATONS AS OF JANUARY 2003

^a Anthracite, bituminous, and lignite are different coal types with decreasing carbon and heat contents.

^b OECD = Organization for Economic Cooperation and Development. Source: Energy Information Agency, International Energy Outlook 2007 (Washington, D.C.: Department of Energy, 2007), available at http://www.eia.doe.gov/oiaf/ieo/index.html.

With the assumption that existing laws and policies do not change, coal consumption in the United States is projected to rise at an annual rate of 1.5 percent, from 24 exajoules in 2004 to 36 exajoules in 2030, when it will provide 57 percent of the electricity produced. Meanwhile, India’s coal consumption is projected to increase from 8.5 to 15.8 exajoules in the same period, again mostly reflecting growth in electricity-generating capacity.

Are these projections meaningful? Probably not. Despite the enormous supply of coal and the growing demand for electricity, other factors will shape future coal consumption. Concerns about climate change and other environmental issues, on the one hand, and consequent changes in government policy, on the other, will surely affect consumption. One day we may see, for example, some sort of cap-and-trade or tax system for carbon that will significantly change the economic equation. These strategies may be important for coal for two reasons. First, there is a great need to increase the efficiencies of coal-fired power plants, particularly in the United States. Second, the climate and environmental concerns will spur development of CO₂ capture and long-term storage.

In the United States, most coal-fired power plants are of the subcritical pulverized coal (PC) type.⁸ In this process, powdered coal and air are injected into the combustion furnace; the heat is used to produce superheated, high-pressure steam; the steam drives a turbine, which produces electricity; and the steam exiting the turbine is condensed to liquid water and returned to the furnace. Along with pollutants (dominantly sulfur and nitrogen oxide gases), the flue gas consists of nitrogen (N₂) from the combustion of air and 10 to 15 percent CO₂. Subcritical PC units have maximum generating efficiencies (the amount of electricity produced per unit of thermal energy in the fuel) of about 34 percent. More efficient are units that use supercritical to ultrasupercritical PC technologies (table 10.3), in which the steam is brought to higher temperatures and pressures and then returned to the furnace as steam rather than as liquid water.⁹ The most advanced designs promise generating efficiencies up to 46 percent, which would represent a decrease in CO₂ emissions of about 20 percent compared with the emissions from subcritical plants.

As an alternative to PC combustion, coal may be burned with limestone in a circulating fluid bed. Although generation efficiency is not much better in this process than for subcritical PC units, circulating-fluid-bed units have the advantage of being suitable for low-grade coals and even peat and waste fuels. More important, with this method the sulfur gases are captured by the limestone,¹⁰ and little or no nitrogen oxide is generated, thereby eliminating most pollutants.

TABLE 10.3 COMPARISON OF PERFORMANCE AND COST OF SOME COAL-FIRED, ELECTRICITY-GENERATING TECHNOLOGIES

^a In units of grams per kilowatt hour.

^b Cost refers to cost of electricity (COE) in cents per kilowatt hour. The COE is the constant dollar electricity price required over the life of the plant to provide for all expenses and debt and bring in an acceptable rate of return to investors.

Source: Massachusetts Institute of Technology, The Future of Coal: Options for a Carbon-Constrained World (Cambridge, Mass.: MIT, 2007), available at http://web.mit.edu/coal/.

Capture of CO₂ is technically viable, but it comes with a cost of a significant decrease in generating efficiency (in addition to the capital cost). The common process involves chemical absorption of flue gas CO₂ into a solution containing amine, a nitrogen-bearing organic compound. The solution must be heated to release the CO₂, which then must be compressed for storage. For subcritical PC units built to incorporate CO₂ capture, generating efficiency is expected to be about 25 percent, but units retrofitted but not originally designed for CO₂ capture will have a much lower efficiency of about 14 percent.

The essential difficulty for efficiency is that the low CO₂ content of the flue gas requires that the CO₂ be concentrated. Two technologies exist to get around this problem. One is to burn the coal in an oxygen-rich gas rather than air in a process known as oxy-fuel PC combustion. The resulting flue gas consists mostly of CO₂ and can be compressed without the need for concentration, but the process has the additional cost of producing oxygen. Oxy-fuel PC combustion technology is in early commercial development, and ultimate efficiency and cost effectiveness have yet to be established.

The second approach is to gasify the coal and remove the CO₂ before burning in a technology known as integrated gas combined cycle (IGCC). In this process, coal is gasified to produce a mixture of hydrogen, carbon monoxide, and CO₂; known as syngas, this mixture is then burned in a gas turbine, and the hot turbine exhaust is used to produce steam that drives a steam turbine (thus the name combined cycle).¹¹

Only a few IGCC plants are in commercial operation in the United States and Europe. The capital costs are quite high, so these plants have been established with the help of government subsidies. Nonetheless, as table 10.3 shows, IGCC has an estimated lower cost with CO₂ capture than other technologies have.

Carbon sequestration is the long-term (thousands of years and longer) storage of captured CO₂ in underground rock or possibly ocean reservoirs.¹² (Reforestation can be considered another form of carbon sequestration.) Carbon sequestration in underground rock reservoirs is technically and economically feasible and in fact is being carried out at several locations now. However, it has yet to be demonstrated at the scale and under the range of conditions needed to have an impact on CO₂ emissions. Given the likely continued importance of coal, the extent to which we shall be able to control CO₂ emissions may well hinge on the success of carbon capture and sequestration.

To emphasize the magnitude of the task and the reason that demonstration at scale is important, consider that sequestering a mere 1 gigaton of CO₂ per year will require capture and storage from five hundred 500-megawatt PC power plants, equivalent to all the plants currently in operation in the United States.¹³ Furthermore, 1 gigaton of CO₂ in the subsurface is equivalent to more than 12.6 billion barrels of oil, or about 40 percent of the oil pumped worldwide every year.

Carbon dioxide must be stored in a liquefied form—that is, as a relatively dense fluid with properties more like liquid than like gas, which is achieved by pressurizing it. Liquefication is important because (1) it minimizes the space required, and (2) the denser the fluid, the less likely it is to escape. The appropriate depth for storage is 800 meters (2,600 feet) and more because at the pressures and temperatures encountered at this depth, CO₂ remains liquefied.¹⁴

There are a number of requirements for CO₂ storage, and perhaps most obvious among them is having the right rock. Most oil and other subterranean fluids are not held in otherwise empty underground caverns. Rather, the reservoirs typically consist of pore spaces and fracture networks in rocks such as sandstones. The higher the porosity and the more fractures there are, the more fluid that the rock reservoir can hold. Another important requirement is that the rock be permeable, which means that the fluid must be able to flow through it. Porosity and permeability typically decrease with depth, so good reservoir rocks cannot be too deep, but they must be deep enough so that the injected CO₂ will be dense.

Even liquefied CO₂ is less dense than water, so another requirement is that there be a cap rock above the reservoir rock to prevent the CO₂ from migrating upward and eventually escaping. Rocks with sufficiently low permeability for this purpose include shale, salt, and anhydrite (calcium sulfate). Fortunately, numerous sedimentary basins around the world offer the right conditions. Sedimentary basins are depressions in the deep, crystalline basement (the foundation of igneous or metamorphic rock on which many sedimentary rock sequences rest) in which thousands of meters of layered sedimentary rock have commonly accumulated. Total national and worldwide storage capacities appear to be enormous, albeit poorly known. For example, the estimates for U.S. storage capacity vary from 2 to 3,700 gigatons of CO₂ and for worldwide storage up to 200,000 gigatons of CO₂.¹⁵

Three types of reservoir rocks appear to hold the most promise for large-scale sequestration.¹⁶ The first is abandoned oil and gas fields. Because the fields already had oil and gas in them, they are already known to be suitable as traps and will not impose any adverse environmental effects. Furthermore, most oil and gas fields have been characterized in great detail; they may already have wells and other supporting infrastructure; and in some fields, injection of CO₂ may lead to enhanced oil or gas recovery. However, these fields are also shot full of holes from exploration and production activities. The holes may or may not be well mapped, and unless they all are properly plugged, there is the potential for catastrophic leaks. A second promising reservoir type is porous sedimentary rocks saturated with salty water (brines), also known as saline formations. Such layers are widespread, and the brines are of no use for agriculture or human consumption. The third option for long-term CO₂ storage is coal seams that are too deep for economic mining.

When CO₂ is injected into a reservoir rock, the fluid pressure around the injection well increases, driving CO₂ into the surrounding rocks (figure 10.5). The CO₂ migrates by displacing some proportion (typically 30 to 60 percent) of whatever fluid is occupying the pore space. As the CO₂ permeates the rock, some (5 to 25 percent) will become trapped in the pore spaces. Carbon dioxide also slowly dissolves in brine, increasing its density. Therefore, CO₂-laden brine will tend to sink and remain permanently underground. Finally, over longer periods of time, CO₂ may react with the rocks to form carbonate minerals, which both immobilize the CO₂ further and seal the rock.

A real-world example illustrates how sequestration works. The Sleipner Project in the Norwegian North Sea is one of several large projects worldwide in which sequestration is being carefully monitored to determine how well it is working (figure 10.6).¹⁷ The Norwegian state oil company, Statoil, recovers natural gas commercially from two gas fields. The gas contains 9 percent CO₂, about 70 percent of which is separated for sequestration. This recovered CO₂ is then injected into the Utsira Formation, a poorly consolidated, brine-saturated sandstone 800 to 1,000 meters (2,600 to 3,300 feet) below the seabed. An 80-meter-thick (260-foot) layer of impermeable shale rests on top of the Utsira Formation and prevents the CO₂ from escaping.

The Sleipner Project commenced in October 1996 with a planned injection of 20 megatons of CO₂. By 2005, nearly 7 megatons of CO₂ had been injected with no evidence of leakage. The expectation is that the CO₂ will eventually dissolve in the saline pore water and sink. There is no shortage of storage capacity. The Utsira Formation is about 250 meters (820 feet) thick and at least 50,000 square kilometers (19,000 square miles) in extent, and the brine-filled pores make up about 35 percent of the formation. The storage capacity is estimated at 600 gigatons of CO₂, or sufficient to hold the CO₂ emissions from all European coal- and gas-fired power stations for centuries.

The Sleipner Project is one of several large projects worldwide in which sequestration of CO₂ is being carefully monitored to determine how it is working. The CO₂ is separated from natural gas and injected into the 250-meter-thick (820-foot) Utsira Formation, a brine-saturated sandstone 800 to 1,000 meters (2,600 to 3,300 feet) below the seabed. An overlying 80-meter-thick (260foot) layer of impermeable shale prevents the CO₂ from rising and escaping. The trapped CO₂ will eventually dissolve in the saline pore water and sink. (After Benson et al. 2005:fig. 5.4)

There is also the possibility of storing CO₂ in the ocean.¹⁸ Ocean storage remains largely a matter of theory, however, because there have been no demonstration projects of any size. Furthermore, some concern has been expressed about potential environmental consequences. For example, CO₂ injection would locally reduce the pH of ocean water, which may have a severe, negative impact on the creatures that live there (chapter 4). At this point, too little is known about ocean sequestration to judge its viability.

To be sure, numerous uncertainties and risks remain regarding sequestration. As noted, there have yet to be any projects of a size necessary to demonstrate that sequestration is viable at a scale necessary to offset emissions. Another uncertainty concerns the possible chemical interactions of CO₂ and the fluids and minerals that make up reservoir rocks, information that is essential for evaluating long-term storage potential. In addition, individual storage sites will have to be extensively evaluated for the possibility of leakage. Finally, there is the question of whether power plants can be located near sequestration sites, avoiding the substantial cost of transporting liquid CO₂ long distances by pipeline. Nonetheless, the existing sequestration projects have worked as hoped, and considering the necessary scale at which sequestration will have to be implemented, technical and economic evaluations have generally been positive.¹⁹

The generation of electricity from nuclear power produces no CO₂ (except for relatively small amounts associated with the mining, transport, and production of uranium fuel and with the management of spent fuel) or air pollution and thus may have a significant impact on emissions. For example, a threefold increase in the current production of nuclear energy to 1,000 gigawatts by 2050 would reduce CO₂ emissions by 2.9 to 6.6 gigatons per year, depending on whether nuclear power plants displace coal- or gas-fired plants.²⁰ Furthermore, nuclear power is a functioning technology that already provides a significant proportion of the world’s electricity. In France, 58 reactors supply 80 percent of the country’s electricity needs. It is no surprise, then, that after more than two decades of dwindling support, nuclear power is enjoying a “renaissance,” or, more precisely, attracting renewed interest as global CO₂ emissions mount.

Significant hurdles exist to the expansion of nuclear power, however: cost, safety, storage of high-level radioactive waste, and proliferation of nuclear weapons. One study argues that given the likely technical and economic challenges necessary to overcome these hurdles, there is little point in developing nuclear power unless it can done in a big way so as to lead to substantial reduction in greenhouse-gas emissions.²¹ Such an undertaking would be massive. Expanding nuclear power from its current worldwide production level of about 350 gigawatts to 1,000 gigawatts by 2050, for example, would require construction of sixteen 1,000-megawatt plants per year for the next 40 years.

To appreciate the issues involved, we must start with the basics of how nuclear power is generated. The production begins with uranium. Nearly all (99.28 percent) of uranium exists as the isotope uranium-238, with most of the remainder being uranium-235 (0.72 percent). Uranium-235 undergoes fission, or the process of splitting the nucleus into smaller particles, by interaction with so-called slow or thermal (low-energy) neutrons. There are several possible reactions in the fission process; one, for example, is

The neutrons produced by uranium-235 decay can be slowed down to become available to split other uranium-235 nuclei. With a “critical mass” of uranium-235, this “chain reaction” will continue until the uranium-235 decays to less than the critical mass. In a light-water reactor, the most common current design for reactors, the critical mass is 3 to 5 percent uranium-235 (for weapons, the critical mass is higher than 90 percent). So to make fuel for a reactor, the proportion of uranium-235 must be increased to above its natural proportion, or “enriched.”

The neutrons from uranium-235 decay also interact with uranium-238 nuclei to produce plutonium-239. This reaction itself does not produce energy, but the fission of plutonium-239 adds to the production of energy. Spent fuel consists of uranium-238, plutonium-239, residual amounts of uranium-235, and a variety of other radioactive elements. Although the waste becomes less radioactive with time, it presents a health hazard for thousands of years.

There are two ways of dealing with the spent fuel. In the open (or once-through) fuel cycle, which is used in the United States, the spent fuel is removed from the reactor and stored for a decade or more in pools of water, which allows the fuel to “cool” as the short-lived radioisotopes decay away.²² The waste is then placed in transportable, air-cooled casks to await disposal in a permanent repository, which has become a controversy of its own. The open fuel cycle is relatively inexpensive and simple. More important, the waste is too radioactive to handle, so it is resistant to theft.

In the closed fuel cycle, which is the strategy adopted by France, Great Britain, Russia, and Japan, the residual uranium-235 and plutonium-239 are separated from the spent fuel. Some is mixed back into new uranium stock and fabricated into fuel for another cycle, and the remainder is fashioned into glass “logs” and, along with unreprocessed spent fuel, placed in interim storage. In general, only one recycle is possible. The closed fuel cycle reduces the amount of new fuel required by about 30 percent, but the reprocessing is expensive to the point that this approach will remain uncompetitive unless the price of uranium skyrockets. Reprocessing plants also represent potential safety hazards because of their large inventories of highly radioactive materials, and the residual plutonium can be redirected to the production of nuclear weapons.

Significant expansion of nuclear energy raises the question of the adequacy of uranium reserves (figure 10.7). At the present rate of consumption, there is probably enough mineable uranium for three or four centuries to come.²³ But whether there is enough for expansion of nuclear energy depends first on the scale of the expansion, which itself depends on the cost of electricity and other factors that are inherently difficult to predict. Second, whether there is enough depends on available supplies, which are unknown but can at least be reasonably estimated.

The chemical formula for metatorbernite, a copper uranium phosphate ore, is Cu(UO₂)₂(PO₄)₂ H₂ O. The sample comes from the copper and uranium Musonoi Mine in Shaba, Zaire, which is famous among mineralogists for the fine specimens of uranium minerals it has produced. This specimen is on display at the American Museum of Natural History, New York. (Photograph by J. Newman, American Museum of Natural History)

One “global growth scenario”²⁴ envisions 1,000 to 1,500 reactors with a total capacity of 1,500 gigawatts providing 25 percent of global electricity needs by 2050. (In comparison, today there are 443 reactors worldwide with a capacity of about 350 gigawatts.)²⁵ The growth scenario would require 9.45 megatons of uranium, and if the capacity remained constant from 2050 to 2100, a total of 24.5 megatons of uranium would be needed for the entire twenty-first century.²⁶

As far as uranium supplies are concerned, the so-called identified resources, which include known deposits as well as deposits that can be inferred to exist based on solid geological evidence, have been estimated at more than 11 megatons. “Undiscovered resources”—meaning those that probably exist, at least based on geological inference, but about which nothing is known—constitute an additional 12 megatons, for a total of about 23 megatons.²⁷ So there appears to be enough uranium for nuclear power to provide 20 to 25 percent of the world’s electricity needs during the twenty-first century.

The practicality of expanding nuclear power hinges on four issues: cost, operational safety, waste storage, and weapons proliferation.²⁸ As far as cost is concerned, improvements must be realized in construction, operation, and maintenance of nuclear power plants to make electricity generated from nuclear power competitive. Although the necessary improvements are “plausible,” nuclear power is unlikely to become truly cost effective unless the social and environmental costs of CO₂ emissions are included in the cost of electricity generation.

Operational safety is a concern not only for reactors, but also for reprocessing plants and systems that handle spent fuel. A related issue is the protection of all these facilities from terrorist attack. The world has seen two serious reactor accidents in the first 50 years of commercial nuclear energy production (at Three Mile Island, Pennsylvania, in 1979, and at Chernobyl, Ukraine, in 1986). Reactor design has advanced significantly, and a safety culture has emerged among plant operators. It is estimated that the likelihood of a reactor accident can be reduced from the historic rate of 1 in 10,000 to 1 in 100,000 reactor-years of operation.²⁹ The safety of reprocessing plants is another matter, however. These plants, as noted, have high inventories of highly radioactive materials and have had a higher frequency of accidents;³⁰ moreover, the risk of terrorist attack involving them has not been adequately evaluated.

Safety also demands a properly trained and qualified workforce to manage, operate, maintain, and even construct power plants. In the developed world, the present workforce is aging and not being replaced, whereas in the developing world an adequately trained, indigenous workforce generally does not exist to begin with. A significant expansion of nuclear energy will require renewed and focused efforts at training.

The safety issues do not appear to be insurmountable,³¹ so they should not be considered an obstacle to large-scale expansion of nuclear energy. The bottom line is public acceptance, which will surely require a well-based perception that nuclear power is safe.

As for waste storage, present plans are for permanent disposal in mined repositories. The idea is to build repositories in areas unlikely to be disturbed for thousands of years by volcanic eruption or earthquakes and where radioactive material, even if it does leak from its containers, will not find its way to the surface. Experts generally (but not universally) agree that geological storage is technically feasible, but after a half century of commercial nuclear power production, no permanent waste facility is in operation anywhere in the world.³² Needless to say, no community wants a repository in its backyard, so the issue in large part involves our ability to develop political and legal mechanisms that will allow repositories to be built. In the growth scenario whereby 20 to 25 percent of the world’s electricity is generated by nuclear power by 2050, a new repository like that at Yucca Mountain, Nevada, would have to come on-line somewhere in the world every three or four years, so the issue is not trivial as far as development effort is concerned.

In dealing with this issue, a possibility is to reduce the required number of new repositories, although that may not completely solve the problem. One way to do so is to store spent fuel on the surface for several decades, as is being done now in lieu of any permanent storage option. Because the spent fuel would be less radioactive and would thus produce less heat, it would require less space in a permanent repository, so fewer repositories would be needed.

Alternative, potentially more acceptable strategies to storage have also been proposed. For example, waste canisters might be placed several kilometers below the surface in boreholes, which would then be plugged by a material such as asphalt or clay. Vast, stable regions of the continents are underlain by dense, crystalline rocks suitable for such storage. Among the attractions of this idea is that even if the fuel canisters were to leak, the waste would likely remain isolated because deep groundwater is generally not plentiful and in any case is isolated from the surface. Also, there are plenty of site possibilities. Disposal facilities might be sited with power plants in relatively remote areas. Suitable disposal sites also exist in the seabed, so one might imagine artificial islands distant from human habitation dedicated to both temporary storage and permanent disposal of waste.

Of all the potential hurdles, perhaps the most complex and worrying one is the threat of nuclear weapons proliferation. It involves two separate concerns. The first is the spread of facilities for uranium enrichment. Most of the nuclear power systems operating today require enriched uranium.³³ This requirement provides a rationale to build enrichment plants for countries such as Iran, but the plants may then be redirected to the production of highly enriched, weapons-grade uranium.

The second concern involves the reprocessing of spent fuel. As noted, reprocessing results in the production of plutonium-239, which can be used for weapons.³⁴ Approximately 200 metric tons of plutonium-239 now exist in Europe, Russia, and Japan, enough for more than 25,000 nuclear weapons.³⁵ One risk is that these existing and expanding inventories are potential targets of theft. Reprocessing might also spread to countries without adequate safeguards against theft or to countries seeking to divert plutonium to weapons production covertly. The technology for reprocessing is no secret because it is published in the open literature, and so far the world has had limited success at curbing the spread of nuclear weapons. It is therefore difficult to imagine that adequate safeguards can be developed under existing nonproliferation mechanisms in today’s fractured world.

Despite this grim outlook, different combinations of fuel cycle and reactor designs may lessen (but not completely eliminate) the proliferation risk and be deployed in one or two decades.³⁶ One approach is to mix uranium- and thorium-bearing fuels.³⁷ Thorium-232 is not fissile, but it does react with slow neutrons to form the fissile isotope uranium-233, which produces energy as well as more neutrons. Proliferation can be thwarted by mixing thorium with uranium fuels in a proportion such that the uranium-233 never reaches more than 20 percent of the mixture. The mixture will consist mainly of uranium-238, so enrichment would be necessary to obtain weapons-grade uranium-233. Enrichment is much more difficult than chemical separation of plutonium-239 from conventional spent fuel. In addition, with this approach much less plutonium-239 is produced than in the standard uranium fuel cycle, and the amount of plutonium-239 as a fraction of all plutonium isotopes is less, rendering the production of weapons-grade plutonium more difficult.

Another approach is to burn thorium fuel containing some plutonium. In this case, fission of plutonium-239 instead of uranium-235 provides the neutrons for the production of uranium-233 from thorium. The proliferation resistance arises from the fact that the spent fuel is highly radioactive and thus very difficult to handle.³⁸ This approach is also a way to get rid of some of the existing plutonium-239 stocks.

Wind and solar power have been called the “iconic renewable resources.”³⁹ The reasons are clear. They do not produce greenhouse gases or otherwise pollute (at least directly); they have the potential to produce vast amounts of electricity forever; they represent a path toward national energy independence and thus security; and they insulate against risks associated with the volatility of fossil-fuel costs. Various analyses suggest that together wind and solar power can practically provide a significant and growing proportion of the world’s primary energy needs through the twenty-first century.

TABLE 10.4 GENERATING COSTS OF WIND AND SOLAR POWER IN 2007 FOR THREE DIFFERENT AMOUNTS OF SUNLIGHT RECEIVED AND WIND VELOCITIES

Source: J. A. Edmonds, M. A. Wise, J. J. Dooley, S. H. Kim, S. J. Smith, P. J. Runci, L. E. Clarke, E. L. Malone, and G. M. Stokes, Wind and Solar Energy: A Core Element of Global Energy Technology Strategy to Address Climate Change (College Park, Md.: Joint Global Change Research Institute, 2007).

Wind power is coming of age now. The cost of producing electricity from large wind power plants is approaching the cost of producing electricity from coal-burning plants, and the readily accessible wind resources are enormous. Meanwhile, the amount of available solar power dwarfs anything else. Capturing solar power with photovoltaic (PV) cells remains relatively expensive, however, and capturing its thermal power with mirrors is also expensive, but has recently become more competitive.⁴⁰ Although the large-scale expansion of solar power lags that of wind power (table 10.4), solar power represents perhaps the greatest hope for meeting future needs while driving down CO₂ emissions.

Nevertheless, both wind and solar power are in their infancy. To emphasize this point, recall that all renewables, which include geothermal and biofuels as well as solar and wind (but exclude hydropower), provided only 2.2 percent of electricity consumed worldwide in 2005. A more specific breakdown of renewable energy resources in the United States in 2006 further illustrates this point (figure 10.8).

The growth of the wind power industry is reflected in the increasing size of wind turbines. Thirty or so years ago a turbine may have had a 10-meter (33-foot) roter (the diameter of the circle traversed by the blade). Today some turbines in the North Sea have 126-meter (413-foot) rotors and generate 5 megawatts of power in 30 mile per hour winds. Wind turbines are grouped together in “wind farms” with typical aggregate capacities of tens to hundreds of megawatts (figure 10.9). As of the beginning of 2008, the world’s largest such facility was the Horse Hollow Wind Energy Center in Texas, where 421 wind turbines generate up to 735 megawatts of electricity, enough to supply the needs of 220,000 homes.⁴¹

The numbers in parentheses are gigawatts of electrical capacity with corresponding percentages of total capacity. Biomass includes wood and derived fuels, landfill gas, municipal solid waste, and agricultural by-products. (Data from Energy Information Agency, http://www.eia.doe.gov/)

The wind farm economies require an average annual wind speed of at least 6 meters per second (13 miles per hour).⁴² The power that wind can theoretically provide is proportional to the cube of its velocity. Thus wind with a speed of 14 miles per hour can provide 25 percent more power than wind moving at 13 miles per hour, although turbines do not capture this entire increase. Consequently, average wind speed is an important factor in determining wind farm location.⁴³

Wind speeds are on average about 90 percent stronger and more consistent over water than over land,⁴⁴ so off shore locations hold enormous potential for development of large-scale wind farms. By one estimate, the available wind off shore between Massachusetts and North Carolina might in principle provide more power than is currently being consumed by this densely populated and energy-intensive part of the world.⁴⁵ The one drawback is that up-front capital costs of off shore development are relatively large. Despite this cost, however, offshore wind farms in Europe now provide more than 1,000 megawatts of electricity annually, and new construction is proceeding rapidly.⁴⁶

Because wind does not blow all the time, it is a variable source of energy. As a practical matter, the variability can be accommodated when wind accounts for less than about 20 percent of the electricity that a power system must deliver in any given hour because the system must be designed with enough flexibility to meet fluctuations in demand anyway.⁴⁷ In systems where wind provides more than about 20 percent of the electricity, a means of storing energy becomes necessary, which affects cost. The example of Denmark, which in 2007 obtained 21.2 percent of its energy from wind power, is telling. One major reason for Denmark’s notable success in the development of wind power is that the country is connected by transmission lines to other countries, which helps to modulate the swings in wind electricity production. Nearly all of Norway’s energy is produced by hydropower, so Norway in particular can absorb the swings by using excess electricity to pump water uphill to high reservoirs and then reusing that water to generate electricity when needed.

Wind power has overwhelming environmental benefits over nuclear and fossil-fuel-burning plants. Besides being nonpolluting, wind power uses almost no water (which is needed only to wash the rotor blades), an increasingly precious commodity required in large quantities in nuclear and coal-fired plants. Concerns for aesthetics and wildlife, in particular the fate of birds and bats, have nonetheless led to local opposition to some wind power projects. Wind turbines can indeed pose a significant danger to birds, the well-publicized example being a wind farm on Altamont Pass, California, home to many raptors. However, a wide-ranging study found that deaths from wind turbines account for only a minuscule fraction of the 500 million to 1 billion bird deaths in the United States annually.⁴⁸ In particular, the study estimated that 72 percent of the fatalities are caused by birds flying into windows, buildings, and power lines. Predation by cats accounts for another 11 percent, automobiles for 9 percent, pesticides for 7 percent, and communications towers for 0.5 percent, whereas wind turbines account for less than 0.01 percent. Avian death rate is site specific and can be minimized by placing wind farms where the impacts on birds will be low.

That wind power is poised for significant expansion is suggested by the increasingly rapid rise of total global installed wind power capacity. Since 1996, that capacity has increased an average of about 28 percent per year to 94.1 gigawatts by the end of 2007 (figure 10.10) and is projected to expand to 240 gigawatts by 2012.⁴⁹ The expansion is being driven by a combination of factors. One, of course, is the world’s insatiable demand for energy, which is driving up the costs of fossil fuels and making wind power increasingly more cost competitive. Another is government policies that encourage its development. Also supporting the expansion is the fact that wind power and turbine manufacturing are increasingly becoming the provenance of large oil, utility, and manufacturing entities that can provide the significant investment necessary to increase capacity and satisfy soaring demand.

Wind power, which in 2008 provided for more than 1 percent of the world's electricity demand, expanded in terms of capacity at more than 25 percent a year in 2006 and 2007, with capacity expected to reach 240 gigawatts by 2012. (Data from Global Wind Energy Council 2007 Annual Report, http://www.gwec.net/index.php?id=90)

The European Union (EU) has traditionally been the largest market for wind power. As of the end of 2007, Germany remained the leader in installed capacity, followed by the United States, Spain, India, and China (table 10.5). Total EU installed capacity that year was 56.5 gigawatts (equivalent to 90 megatons of CO₂ emissions), or 3.8 percent of total electricity demand, up from 0.9 percent in 2000.⁵⁰ The EU has adopted the goal of obtaining 20 percent of its energy from renewables by 2020. If this goal is to be met, about 12 percent of all electricity will have to come from wind power.⁵¹ Reaching the goal will require massive development, especially of off shore wind resources.

In 2007, installed wind power capacity in the United States increased 45 percent over the previous year, to 16.8 gigawatts.⁵² The expansion has been encouraged by the 2 cents per kilowatt hour federal production tax credit. A high growth rate is likely to be sustained because 24 states have adopted the Renewable Portfolio Standards, a mandate that certain proportions of energy must come from renewables by specified dates. Although wind power now satisfies only about 1 percent of electricity demand in the United States, the potential for growth is enormous. In fact, by one scenario (albeit a highly uncertain one because of unknown economic and technological variables), U.S. wind resources may conceivably expand to make up about 25 percent of electricity output by 2030.⁵³

TABLE 10.5 NATIONAL INSTALLED WIND POWER CAPACITIES AS OF THE END OF 2007

Note: Data are for countries with capacities greater than 1,000 megawatts.

^a For 2006.

^b As a proportion of total national electricity generation.

Sources: Global Wind Energy Council, Global Wind 2007 Report (Brussels: Global Wind Energy Council, 2007), available at http://www.gwec.net/index.php?id=90; British Petroleum, BP Statistical Review of World Energy 2007 (London: British Petroleum, 2007), available at http://www.bp.com/productlanding.do?categoryId=6848&contentId=7033471.

Several hurdles to the massive expansion of wind power in the United States remain, however.⁵⁴ The most important is that the existing transmission infrastructure is inadequate. It is particularly a problem in the upper Midwest, where wind is a plentiful but “stranded” resource because the existing transmission system is not capable of transporting large amounts of power from potential generation sites to population centers. Another hurdle is energy storage, which, as suggested earlier, practically limits the proportion of power that wind can contribute to a local power grid.

The technology to turn the Sun’s energy into electricity is hardly new. Spurred on by the oil crisis of the 1970s, nine solar thermal power plants with an aggregate capacity of 354 megawatts were constructed in the 1980s in the Mojave Desert, California. Interest in solar power waned as oil prices dropped, only to reawaken again as prices began to rise after 2002. In the meantime, it has gained a foothold elsewhere, especially in Japan and Germany.

There are two ways of converting solar radiation to electricity. One is by transforming solar radiation directly into electricity using PV cells; the other is to generate electricity by collecting solar heat with the use of mirrors. PV cells⁵⁵ utilize the photoelectric effect, the process of ejecting an electron from an atom by the atom’s absorption of an incident photon, or a packet (quantum) of light. In its simplest form, the PV cell consists of layers of p and n semiconductor material, most commonly silicon, sandwiched between layers of conducting materials. Sunlight falling on the n layer produces electrons, which are collected in the p layer to create a voltage difference between the front and back faces of the cell. The voltage difference drives an electric current when the two sides are connected by a conductor. The amount of electricity produced by a PV cell is approximately proportional to the amount of solar radiation falling on it. PV cells are mounted into modules (40 cells are common), which are in turn combined to form arrays of various sizes depending on application (figure 10.11).

PV cells offer great flexibility. For example, arrays can be mounted on residential and commercial buildings. This configuration works especially well if the electricity produced locally is tied into the electrical grid, which obviates the need for expensive storage devices, such as banks of batteries, and excess energy can be sold or traded for electricity when none is to be had from the PV arrays. PV arrays are also suitable for driving devices isolated from other sources of power, such as water pumps on remote farm pastures.

In 2007, the cost of electricity from PV cells was more than 20 cents per kilowatt hour, depending on the intensity of sunlight (see table 10.4), whereas electricity from the burning of coal cost only 2 to 4 cents per kilowatt hour. Great effort, therefore, is being directed at producing PV cells with greater efficiencies at converting solar energy to electricity. Solar radiation consists of photons with a spectrum of energies, but specific semiconductor materials can absorb photons only in narrow ranges of energies, limiting their efficiencies. Therefore, one way to gain efficiency is to construct cells consisting of two or three layers of different semiconductor materials that absorb photons in different parts of the energy spectrum. The typical efficiency of commercially available PV cells is about 3 to 14 percent, but efficiencies up to 50 percent are theoretically possible. Development of more efficient PV cells and less expensive ways to manufacture them are at the technological forefront and promise substantial improvements in costs.

As noted, electricity may also be generated from solar energy by using mirrors to collect the heat. This process is practical in cloud-free regions and best in those regions that receive more than 6.7 watts of solar radiation per square meter, such as in most deserts of the world and countries with the right geographical and climatic characteristics, such as Spain. There are three approaches to “concentrating solar power,” as the process is known. The first is to use a linear array of parabolic troughs to focus the light onto an oil- or molten salt–filled pipe running down the center of the trough (figure 10.12). The troughs can be oriented in a north–south direction and track the sun as the day progresses. A second approach is to use a large field of sun-tracking mirrors to focus sunlight onto a receiver at the top of a central tower, where the focused light heats oil or molten salt. In either case, the hot fluid is used to boil water, which in turn drives a steam turbine.

Molten salt can also be used to store heat to provide a continuous flow of electricity at night or during cloudy days. The AndaSol project, a 150-megawatt solar thermal plant in the province of Granada, Spain, uses molten salt storage.⁵⁶ Parabolic troughs heat molten nitrate salts. This fluid is pumped into a hot storage tank at 386°C (727°F) and then run through a heat exchanger to boil water as it is transferred to a “cool” tank at 292°C (558°F). The system maintains more than seven hours of storage capacity at full output.

Parabolic mirrors focus sunlight on a fluid-filled pipe and track the sun at a plant located near Kramer Junction, California. This is one of a number of such plants with a combined capacity of 354 megawatts built in the 1980s after the first oil crisis. (Department of Energy, National Renewable Laboratory, http://www.nrel.gov/data/pix/, photograph 15236 by Warren Gretz, with permission).

A third approach to concentrating solar power uses an array of mirrors mounted on a sun-tracking dish. The mirrors focus the sunlight onto a “receiver,” which collects and transfers the heat to a Stirling engine.⁵⁷ The conceptual basis of the Stirling engine dates back nearly 200 years to 1816, when the Scottish inventor Rev. Dr. Robert Stirling (1790–1878)⁵⁸ conceived of the “heat economizer,” a more fuel-efficient engine to replace the steam boiler. The engine works by the heating and cooling of a gas (hydrogen in the modern engine) in a sealed cylinder. As the gas expands and contracts with changing temperature, it drives the pistons of a four-cylinder reciprocating engine, which in turn drives a generator. The system is extremely efficient, having achieved a system-to-grid conversion efficiency of more than 30 percent. Two generating stations based on the Stirling engine are now under construction in southern California. Their combined initial generating capacity will be 800 megawatts, making them among the largest solar power stations to date.

Like wind power, solar power is now expanding at an increasingly rapid rate. Germany, a country with only modest solar resources, accounts for much of the accelerating growth (figure 10.13). The reason for the growth there is that the government extended its “feed-in tariff” program, originally established for wind power, to all PV producers of electricity, large and small alike. Feed-in tariff programs ensure a producer of being able to connect its local system to the power grid and either sell excess power to the grid operator or trade it for electricity delivered at another time. The policy typically drives a modest increase in electricity cost because the grid operator must buy solar-generated electricity at a price above which it can otherwise produce electricity. In the case of Germany, the premium is substantial and guaranteed for a number of years. As a result, it suddenly became profitable for German home and commercial building owners to install solar panels. The policy has been wildly successful at forcing greater PV market penetration, and it has thrust Germany into a leading position as a developer of solar power technology.⁵⁹ The idea is catching on and slowly spreading beyond Europe.

Efficiency dictates that solar power plants be located in sun-soaked parts of the world, so the development of a network of high-voltage transmission lines is important to the large-scale expansion of solar power. (Wind power would also benefit from such a network, but the problem is not so severe because wind power facilities can more commonly be built closer to population centers.) Nearly all domestic and industrial equipment uses alternating current (AC), so nearly all transmission lines are AC as well (98 percent by distance in United States). High-voltage AC (HVAC) transmission lines lose energy, depending on length of transmission, however. The alternative is direct current (DC) lines, which lose less energy, are more stable, and have several other advantages compared with HVAC lines. But HVDC lines require expensive converters from AC to DC and back again on the ends of the line. Consequently, HVAC lines are cheaper for distances of less than 500 to 800 kilometers (300 to 500 miles), whereas HVDC lines are more economical at greater distances.⁶⁰ The expense of converters also means that HVDC lines are not suited to providing electricity at points along a route because each point would require another converter. A logical strategy is to develop a long-distance backbone of HVDC transmission cables that feed into local and regional HVAC networks.

Several grand ideas for expanding solar power have been proposed. One argues for the construction of 120,000 square kilometers (46,000 square miles) of solar installations in the southwestern United States with an accompanying HVDC transmission network.⁶¹ The cost would be about $400 billion over the next 40 years, but the program would provide nearly 70 percent of electricity and more than 33 percent of total energy requirements by 2050. In Europe, a plan known as Desertec has been put forward.⁶² The plan is to harvest solar and wind energy with plants situated throughout North Africa and the Middle East and to distribute the energy via a “megagrid” of HVDC transmission lines connecting all of Europe, North Africa, and most of the Middle East. This ambitious plan would cost approximately €400 billion and supply about 17 percent of Europe’s electrical needs by 2050. It involves numerous unanswered questions, however, including how to gauge risks from terrorism and political instability.

Providing for the world’s growing energy needs and simultaneously gaining control of and then decreasing CO₂ emissions are essential if we are to avoid potentially catastrophic climate change. We already know how to reduce emissions—the knowledge and nearly all the technology needed to achieve this goal are at hand. Yet reducing emissions to 2000 levels, the illustrative but somewhat arbitrary goal put forth at the beginning of this chapter, remains a daunting task.

To appreciate just how daunting the task is, think back to the challenges of producing clean electricity. Recall that sequestration of 1 gigaton of CO₂ per year from traditional coal-fired power plants will require capture and storage from five hundred 500-megawatt plants, and that 1 gigaton of CO₂ in the subsurface is equivalent to about 40 percent of the oil pumped worldwide every year. Recall that expanding nuclear power to 1,000 gigawatts by 2050 will require construction of sixteen 1,000-megawatt plants per year for the next 40 years and bringing on-line a large waste repository every several years. And recall that although both wind and solar power capabilities are growing fast, they still constitute but a minuscule fraction of worldwide electricity generation. Implementing solutions on the scale necessary to reduce CO₂ emissions to levels below those in 2000 will probably require the development of programs that have the scale and urgency akin to the Apollo Space Program or its mobilization at the start of World War II.⁶³ Furthermore, these efforts have to be global, led in particular by China, the United States, and Europe.

The shift from dirty to clean electricity will take decades to implement. In the meantime, other steps not covered in this book will be essential. Most important will be the rapid implementation of energy conservation and efficiency measures to get us through the next decade or two. These measures include moving to hybrid and electric vehicles; finding ways to reduce vehicle travel distances by expanding mass transport; retrofitting existing buildings and constructing new ones with efficient insulation, space heating, cooling, and lighting; and fitting individual buildings with PV panels to capture solar energy or devices to capture geothermal energy. Another is to substitute natural gas for coal to the maximum extent possible as the fuel for electricity production. And it is not out of the question to hope that the next decade will see the implementation of promising new technologies, such as those that will directly remove CO₂ from the air.⁶⁴

Each stabilization wedge represents an action that reduces the CO₂ emissions rate. The actions are deployed gradually, but at an increasing rate beginning in 2008. They combine to initially maintain emissions constant at the 2008 rate, but at some point result in a decrease in emissions rate with time as new technologies are deployed.

At the current time, however, there is simply no silver bullet to getting rid of the problem of greenhouse-gas emissions—efforts on a number of fronts are necessary. Various stabilization scenarios, or models of energy and economic development along with the emissions they would create, have been developed to explore the mitigation options. These scenarios assume that mitigation activities begin gradually in the near future and increase with time. A useful and widely cited conceptualization of this notion illustrates how mitigation efforts can be practically combined to achieve the desired reduction.

The conceptualization begins with a “stabilization triangle,”⁶⁵ one leg of which is the growth of CO₂ emissions projected out for the next 50 years at a business-as-usual rate, and the other leg of which represents an initially constant but then decreasing emissions rate (figure 10.14). The stabilization triangle may be divided into several “stabilization wedges,” each of which represents an action that reduces the emission rate by an equal amount. The notion of a wedge implies that the action begins now and increases with time. In addition to the emissions-reducing strategies associated with energy production, possible wedges include accelerating measures to increase energy efficiency and conservation, reducing deforestation in combination with replanting campaigns, and adopting agricultural practices that limit the loss of organic material from soils by controlling erosion and practicing conservation tillage (the practice of planting seeds in boreholes in soil rather than by plowing).⁶⁶

An increase of 1.5 percent per year in the current annual anthropogenic emission rate of 36 gigatons of CO₂ per year projects to a 2058 emission rate of 76 gigatons of CO₂ per year. (The 1.5 percent figure is the historical growth rate, but the rate of increase in CO₂ emissions is now depressingly closer to 3 percent.) Stabilization just to the 2008 emission rate thus would require mitigation actions that save 40 gigatons a year. One can argue over which mitigation actions to take and how extensive they should be, but the point of the wedge concept is that it offers a way of thinking about practical solutions to a problem that may otherwise seem intractable due to its enormity.

This book is about how the climate system works, but it ends by emphasizing how tightly climate change is entwined with energy production. Energy production, in turn, has become a bellwether of economic development. The connection illustrates how civilization is now leading us down a path never before trodden, at least by our species. For the first time, humanity can affect the climate of the entire planet and the vast biogeochemical cycles that control the character of its surface. In a way, it is astonishing that we even know about such matters as these cycles because their scales of space and time are so far beyond those of our own lives. Think back 170 years ago to when Louis Agassiz began to understand that his world had once been covered with ice. At the time, there was hardly a conception of climate, let alone climate change, and humans were passive riders on Earth. Then think about where we are now as we peer at our planet from above with satellites, explore the deep ocean with robots, and delve into Earth’s past by drilling into deep layers of rock and ice, groping for an understanding of the intricate interactions of the Sun, atmosphere, ocean, ice, life, and land—as well as of the changes forced by human activities.

And then think further: What if we had not learned the things we now know about climate and the possible dangers to civilization that climate change poses? This knowledge attests to our intelligence and innovation, just those characteristics that got us into the present predicament. We have to believe that they will also get us out of it.