The Physics of Energy, page 162
Table 36.3 Characteristics of existing and potential low-carbon energy resources. Data on fossil fuels are included for comparison. Resource base estimates are rough order-of-magnitude estimates only. Where question marks appear, they indicate quantities that are impossible to estimate at this time. High-quality resources produce mechanical energy directly or thermal energy at very high temperature; medium-quality resources produce thermal energy at temperatures in the range300–1000; low-quality resources exploit temperature differences of 30∘C or less.
1 Deployable with existing technology
2 Demonstration technology exists, but technical problems, perhaps major, make deployment impractical at present
3 Facing technical challenges of unknown difficulty that may never be overcome
Order-of-magnitude accuracy unless more precision is shown
System for harvesting energy scales with size as
Over land between latitudes and
Output of largest existing facility
Effectively a form of solar power
From [282].
36.6.2 Large-scale Non-fossil Energy Resources
Focusing on the largest-scale non-fossil energy resources, we summarize the situation for solar, wind, ocean thermal energy, geothermal energy, and biofuels. As can be seen from the figure in Box 36.3, where these resources are found at the top of the graph, and from Table 36.3, these are the only sources among those studied in this book that have sufficient total power available to in principle power human civilization in perpetuity (or for thousands of years in the case of geothermal). Only solar and wind power could be harvested using existing technologies at the scale required to replace a significant fraction of the roughly 15 TW of energy used by humanity in 2016 or the 30 TW that may be needed within a few decades if current trends continue. The technology for exploiting enhanced geothermal systems has not yet reached the level of maturity of solar and wind energy systems, and the exploration of energy production by biological systems in ways that approach the efficiency of existing solar systems is in its infancy; these energy sources could in principle provide terawatts of power, however, in the intermediate-term or long-term future.
Solar power (§22, §25) Earth receives solar energy at a rate of 173 000 TW at the top of the atmosphere. This averages to 343 W/m2 over Earth’s surface. After reductions due to atmospheric absorption and scattering, including weather induced intermittency, about 190 W/m2 (or ~180 W/m2 of exergy) is available on average for capture. Solar power is a two-dimensional resource (Box 36.3); the recoverable energy scales with the area of the devices that are deployed to harvest the power. Existing solar power extraction systems, however, can capture only a small fraction of the exergy incident on a given area of solar collectors. Current solar thermal energy systems can convert only a net 3–5% of incident solar energy to useful electrical power over large areas of land. Using currently economical technology, large-scale power stations using photovoltaics can achieve gross efficiencies of around 6%. Even so, as discussed in the solar energy chapters, much of human energy needs could be supplied by covering a small fraction (1% or less per TW) of the world’s desert areas and/or urban areas with solar power systems. While the most immediate potential use of solar electric power is to supply energy to the electrical grid, or to supply power for individual residences or commercial buildings, more generally solar electric power could be used to power land transport and heat pumps for residential and commercial heating, so that solar electric power could be used to cover a broad spectrum of human energy uses; this and related issues are discussed further in §36.6.4.
Wind (§28) Winds are driven by temperature and pressure gradients that originate in the uneven pattern of solar energy absorption over Earth’s surface. Of the 120 000 TW of solar power absorbed by Earth (assuming an albedo of 0.3), between 1/2% and 3% is believed to be fed into winds. Of this ~1000 TW, estimates of the power that can be extracted from the wind for human use range between ~1 and ~70 TW. (The uncertainty in these estimates is discussed at length in §28.3.2.)
Although wind energy is distributed over Earth’s entire surface, a wind harvesting system cannot be extended over an arbitrarily large area. On the one hand, wind energy can be extracted from more than one single line of turbines placed along a direction perpendicular to the wind, as more energy mixes down from above after a distance of several times the turbine height. This general idea governs the optimal placement of wind turbines in the downstream direction as discussed in §28. On the other hand, placing wind turbines over an area many hundreds of kilometers in each direction will have the potential to deplete the wind up to near the top of the troposphere, so that downstream turbines will not have power. Uncertainty in the way that the finite vertical nature of the resource impacts power captured by large-scale wind harvesting systems is one reason for the wide range in the estimates of total power that can be extracted from the wind. A consequence of the spacing needs between wind turbines is that land use (measured in km2/MW) is higher for wind turbines than for solar plants. Wind turbines, however, can more easily be combined with other uses of land than can solar installations.
Wind turbine technology has reached a state where wind farms placed in favorable locations could likely supply a substantial fraction of current electric power needs. A primary technological challenge facing wind (and solar) electric power is the need for large-scale storage (§37) to better match the supply of renewable electric power with demand (§38).
Geothermal (§32) While standard hydrothermal systems probably cannot ever replace a substantial fraction of human energy use, geothermal energy available in deep hot dry rock could in principle power humanity for thousands or tens of thousands of years. The technology necessary to effectively extract this enhanced geothermal energy is not yet in place, though it is not far beyond the current range of closely related drilling and extraction technologies developed in the oil industry. If concerns about seismic impact of deep drilling and fracking can be ameliorated, if underground fluid losses can be controlled, and if efficient and cost-effective systems can be developed, in principle enhanced geothermal systems could provide a non-renewable replacement for fossil fuels that would last for many hundreds or thousands of years.
Biofuels (§26) In discussing potential large-scale energy resources that may eventually replace fossil fuels, energy from biological systems cannot be neglected. A present limitation is that the total rate of conversion of solar energy to terrestrial biomass is only a factor of three or so greater than total current human energy use. A large fraction of biomass is not easily accessible to human use, and much land is already used for growing crops needed for food. Thus, existing biofuels cannot be scaled up to provide a substantial fraction of human energy use. On the other hand, the study of biological energy conversion, and biosystems in general, is at a very early stage of development; in principle biological systems working at the molecular level have much finer local control over solar energy conversion processes than simpler homogeneous materials like a silicon cell. One can imagine hybrid semi-biological systems that combine relatives of photosynthesizing organic material with an engineered physical substrate to provide cheap but highly efficient self-reproducing systems that capture solar energy more effectively than current photovoltaic or solar thermal systems. In a general sense, biofuels can be thought of as solar energy capture devices, a category that also contains photovoltaics and solar thermal power plants. At present, the best biofuels are about a factor of 100 less efficient than a photovoltaic array as a source of electricity (§26.5). Thus several orders of magnitude increase in efficiency would be necessary for advanced biological systems to compete with solar energy as a source of electricity. Nevertheless the ultimate potential of biofuels is unknown and their potential to provide liquid hydrocarbons directly from a renewable source should not be ignored.
Ocean thermal (OTEC) (§27) Ocean thermal energy conversion seeks to use the temperature difference between the surface waters and water at depth in tropical oceans to drive a heat engine. As such, OTEC is clearly a two-dimensional resource, spread around Earth’s equatorial zone. A reasonable upper bound (§27.6) on the energy density that could be available for exploitation by OTEC is given by the rate at which solar energy is absorbed at the ocean’s surface, roughly 30 W/m2. Estimating the area of near-equatorial waters to be km2, we find a rough estimate of the total OTEC resource to be 4000 TW, corresponding to the rate of ocean thermal energy transport away from the equatorial region; this total renewable resource power is exceeded only by solar power itself.
Unfortunately, ocean thermal energy is a low-quality resource (in the technical sense, i.e. (exergy/energy)) because the temperature differences involved are no more than 17–22. The Carnot limit on the 1st law efficiency of an OTEC system is therefore in the range of 5–7%, and the exergy available is on the order of 200 TW, or roughly 2 W/m2. Given irreversibilities and other losses, 2nd law efficiencies are also relatively low. While the vast extent of the resource has sustained interest in further developing OTEC systems, the low exergy density, which is roughly two orders of magnitude smaller than the original solar resource, makes it clear that terawatt-scale utilization of OTEC would be impossible without covering a sizable fraction of the ocean’s surface with low-efficiency energy capture devices. So far OTEC installations have been limited to research and demonstration facilities.
In summary, even if fossil fuels were to completely vanish from the planet over the next 20 years, humanity would have the capacity to replace fossil power sources through solar, wind, and other resources. Technical issues such as storage and transmission remain, but the primary obstacle to such a transition currently is economic. Because of their high energy density and easy extraction, fossil fuels are still the least expensive energy resource for most purposes, though solar and wind energy systems are rapidly becoming competitive in many markets. It seems inevitable that over time fossil fuel energy costs will increase as supply and accessibility decrease, while solar energy, in particular, is likely to continue to decrease in cost over the coming decades. Once solar or wind (or geothermal) energy becomes cheaper on a per watt basis than fossil fuels in a sufficiently wide range of contexts, society may begin a rapid transformation away from fossil fuels to non-carbon based energy resources.
36.6.3 Resources with High Energy Density
For replacement of a large fraction of current fossil fuel based energy sources on the scale of many terawatts in the near term the only viable options are large-scale resources such as solar and wind, and potentially geothermal. These large-scale resources are, however, generally quite diffuse. From the economic point of view, it is desirable to identify carbon-free energy resources with high power density. A glance at Table 36.3 and the right-hand side of the figure in Box 36.3 identifies hydropower, wave power, and marine and tidal current power in addition to nuclear energy as such resources. Even though some of these resources are not sufficiently extensive to produce many terawatts of power, it may be economically advantageous to exploit them in favorable locations. Even producing a fraction of a terawatt with each of these resources could help provide a substantial quantity of high-quality carbon-free electrical power in an economically optimal fashion. We briefly comment on each of these resource options.
Hydropower (§31.1) Even though the total resource base is at most a few terawatts, three characteristics make hydropower a very attractive high-power-density resource. First, hydropower can be concentrated in small areas that are essentially pointlike on the terrestrial scale. Hydroelectric generating facilities that are capable of producing energy at rates exceeding 10 GW can therefore be relatively compact, although the area needed for the associated reservoirs makes hydropower less efficient in overall land use than solar power. Second, the resource has high quality (exergy/energy) so conversion to electricity can be very efficient. And third, the resource takes a simple mechanical form – gravitational potential energy of stored water – which can be converted to electricity with well-understood technologies. Hydropower is currently the only renewable energy source that contributes substantially (~3%) to world energy use, providing roughly 0.43 TW of electrical power in 2013. It has been estimated that this contribution could be increased to 1 TW or more over the next century.
Wave energy (§31.2) Although ocean wave energy is distributed over the surface of the world’s oceans, as mentioned in Box 36.3, wave power should be considered a one-dimensional resource. For example, placing a line of wave energy extracting devices running north–south in an area where waves move in an eastward direction could extract all wave energy over a substantial surface area of ocean.
Annual average wave power near oceanic coastlines ranges from roughly 10 kW/m to over 60 kW/m. Assessments of the total wave power resource are around 2–3 TW, though for reasons discussed in §31.2.4, only a small fraction of this would likely ever be accessible to human use. Nevertheless, the high energy density may make this resource economically favorable in some locations as technologies for wave energy capture develop.
Tidal and marine current energy (§31.3, §31.4) The tides dissipate about 4 TW continuously, mostly through friction with the ocean floor. Only a small fraction of tidal power, perhaps 60–120 GW, is available for human utilization through tidal barrages (§31.3.3) or turbines placed in tidal streams (§31.3.4). The energy density and scaling properties of tidal barrages are similar to hydropower; since there are fewer good locations and since tidal barrages have lower hydraulic heads than dams used for hydropower, their potential as energy sources is a small fraction of that of hydropower. In favorable locations, however, existing tidal barrages are capable of generating as much as 250 MW at peak power.
Because water is almost 1000 times as dense as air, moving water carries much more energy than air moving at a similar speed. Where tidal flows are constrained to narrow streams, underwater turbines can harvest significant amounts of power. Power densities in favorable tidal locations are greater than 10 kW/m2. In some ocean locations, large-scale marine currents can give similar power density. Despite the logistical challenges, the high power density may make devices such as the tidal stream underwater turbines described in §31.3 economically favorable in locations with substantial tidal flow and appropriate coastline geography. Marine currents such as the Gulf Stream may also be a source of relatively dense power. Like tidal power, total marine current potential is also substantially less than one terawatt.
Nuclear fission power (§16) Nuclear power, while not renewable, is also carbon-free. The attraction of nuclear power is the extremely high energy density of this resource. In §16 we concluded that roughly 200 tonnes of natural uranium could fuel a 1 GWe (assumed ~3 GWth) once-through fission reactor for a year. This corresponds to an energy density of about 500 GJ/kg. So the energy content of a single kilogram of natural uranium is equal to the energy content of about 15 000 L of gasoline.
Uranium is widely distributed over Earth’s crust at an abundance of about 2.7 ppmm, so Earth’s crust can be viewed as an energy resource with about 1.3 MJ/kg due to its uranium content, although the cost to extract uranium at this concentration is prohibitive. Like fossil fuels, uranium for nuclear fission power is a resource that has been “integrated over time,” forming areas of high energy density that are pointlike (zero dimensional) on a terrestrial scale. At present, economically viable uranium deposits have about 0.1% uranium on average, about 400 times the average crustal abundance.
As discussed in detail in §16, world uranium resources recoverable at a cost of $ 260/kg U (see [76] for details) are currently estimated at t. With a total energy content of about 3.6 ZJ, this uranium could provide about 270 GWe of nuclear electrical power (the present value) from once-through reactors for about 150 years (§16.2.2). While the cost of uranium is not a major component of the cost of a nuclear fission reactor, and economically recoverable reserves could grow significantly with increased demand, expanding the contribution of nuclear fission power to electric power production beyond that of hydropower would be difficult to achieve with once-through fission reactors, and impossible to sustain over the long term. Much more energy could become available, however, if breeder reactors that produce fission fuel from the common isotope of uranium () or from thorium can be made safe and economically viable.
The high energy density of nuclear fission power is offset by the technical complexity of nuclear power plants and by externalities including the environmental impact of uranium mining and enrichment, the hazard presented by radioactive waste, and the threat of nuclear proliferation. The extent to which nuclear energy will be used as part of the world’s energy portfolio in the twenty-first century will depend to a large extent upon how the risk of nuclear power is weighed by society relative to the risks and associated costs of fossil fuels and other energy options.
Nuclear fusion power (§19) Nuclear fusion power based on deuterium–tritium (dt) fusion resembles fossil fuels and nuclear fission power in that it is a high-energy-density resource that relies on a naturally occurring substance (in this case lithium) that has been concentrated into enriched ores over geological time. Nuclear fusion, however, differs in one critical way: the technology to exploit nuclear fusion does not yet exist. As explained in §18.4.3, only dt fusion is under consideration at the present time. Tritium, however, does not occur naturally on Earth; it must be created by bombarding lithium with neutrons. Each atom of lithium yields one atom of tritium, which fused with deuterium gives off 17.6 MeV of energy. This corresponds to a potential energy density of about 220 TJ/kg of lithium (Problem 36.25), higher even than that of uranium. Lithium, like uranium, is in limited supply. According to [25] a rough estimate of the world’s lithium resources in 2016 was 40 Mt, corresponding to a total fusion energy potential of about ~10 YJ or approximately TWy. If even a small fraction of this energy could be captured by a practical fusion reactor, it is enough to provide 30 TW of power to humanity for hundreds if not thousands of years. If, eventually, dd fusion is developed, then the abundance of deuterium (1o of the hydrogen in Earth’s oceans) would assure an adequate energy supply for an even longer time.
