Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
R&D Challenges in the Chemical Sciences to Enable Widespread Utilization of Renewable Energy Nathan S. Lewis, California Institute of Technology When looking at the present utilization of primary power and options for tuture sources of energy, a number of questions can be envisioned. Where do we presently get our power from and what are the costs? What is the role for the chemical sciences in renewable energy technology? To answer these questions, it is useful to review the present primary power mix how much energy is con- sumed, from what sources, future constraints imposed by sustainability, and the theoretical and practical energy potential of various renewables. Once these issues have been addressed, it is possible to identify the challenges for the chemical sciences to economically exploit renewables on a scale commensurate with our energy needs. The mean global energy consumption rate in 1998 amounted to 12.8 TW (383 quad/year, of which only 10 percent was used for electricity (Figure 5.1~. Oil, gas, and coal constitute almost 80 percent of this total energy consumption. The mean U.S. energy consumption rate in 1998 was 3.3 TW (99 quad/year), 15 percent of which was used to generate electricity. At present, the use of renewables for energy accounts for only a small per- centage of total energy requirements. Of the 12.8 TW of global power consumed in 1998, power from biomass accounted for only 1.21 TW, while the largest renewable source consumed hydroelectric still accounted for only 0.3 TW. One possible explanation for the limited use of renewables may be cost. With one exception, the use of fossil fuels for electricity generation in the United States is presently by far the least expensive option. The cost per kilowatt-hour for il quad = 1015 Btu = 1.055 x 10~8 I; ~ quad/year = 0.0334 TW. 33
34 TW Os~0 hi' ^ an' FIGURE S.1 Mean global energy consumption, 1998. ENERGY AND TRANSPORTATION 2.96 2.7 4~ '''1 5- , 4 - 3 - 2 - O 0> ~ ~ )~ 0.286 0.286 0.828 a;\ c\ electricity derived from coal is 2.1 cents; for natural gas is 3.6 cents; and for oil, 3.9 cents. This compares with a cost per kilowatt-hour for wind of 3 to 5 cents, and for solar approximately 22 cents. However, it must be noted that electricity derived from nuclear power is quite competitive with fossil fuels, with a cost of 2.3 cents per kilowatt-hour.2 3 Proven reserves of oil, natural gas, and coal tend to underscore that fossil fuels will remain an abundant inexpensive resource base for the foreseeable future. Proven oil reserves are expected to last at least 40 years, natural gas should last at least 70 years, and coal reserves are adequate for 200 years. By adding reserves that are likely to be found, the oil supply would then last between 50 and 100 years, natural gas 90 to 275 years, and coal at least 2,000 years. In light of this supply, it is reasonable to conclude that renewables will not play a large role in primary power generation unless one of two things happens. Either technology breakthroughs on these renewable sources of energy reduce the cost of these sources significantly below what they are today, or there are unpriced externalities introduced that significantly increase the cost of fossil fuels. For example, environmental concerns may result in the introduction of carbon taxes or subsidies for carbon-neutral technology. Looking to the future, it may be possible to make some predictions regarding primary power demands and the expected environmental impact over the next 50 2These costs represent the costs of capital equipment and fuel but not waste disposal or environ- mental remediation. 3Electricity, however, is high-value energy, and the cost per Joule of energy produced by consum- ing fossil fuels to make heat is approximately a factor of 5 to 10 times less expensive.
R&D CHALLENGES IN THE CHEMICAL SCIENCES 35 to 100 years based on demographics. The global population is expected to rise to 10 billion-11 billion people by 2050, and GDP growth is expected to rise an average of 1.6 percent a year, which is its historical average. Balanced against these factors, energy consumption per unit of gross domestic product (GDP) is expected to decline 1 percent a year, mainly as a result of increases in energy efficiency in the industrialized world. Based on these projections, total primary power consumption is expected to rise from 12 TW in 1990 up to 28 TW by 2050. The overriding question is from where this 28 TW of power will be derived. The above projection, as previously stated, depends on increased efficiency in energy usage. Sustaining the historical trend of carbon intensity in the energy mix implies that by 2050 the energy economy will actually be more efficient than one run entirely on natural gas. This can only be accomplished if there are signifi- cant contributions from carbon-neutral power. Based on projected carbon dioxide emissions from fossil fuel usage, it is estimated that even with increased use of fossil fuels over the coming century, these fuels will not be enough to meet the expected 28 TW of power demanded by 2050. There will be an estimated 10 TW shortfall that must be derived from carbon-neutral sources in order to meet the primary power needs of the planet. Obviously, this is not an insignificant amount of power. Ten terawatts was the entire global power production from all sources in 1990. In addition, if atmo- spheric carbon dioxide emissions are to be stabilized, there will be an even greater need for renewable energy. For example, if man-made releases of carbon dioxide are to be stabilized at 550 ppm twice the preindustrial level of carbon dioxide in the atmosphere 20 to 30 TW of carbon-neutral primary power would be needed.4 This is a daunting amount of power. Without economic policy incentives, the needed technology to meet these demands will probably not be in place soon enough to meet this demand by 2050. In fact, meeting the goal of commercial carbon-neutral power capable of producing 10 to 20 TW by the mid-21st century could require efforts comparable to the Manhattan Project or the Apollo space program. Where such large amounts of power can be derived from must be examined. If this power is to be carbon neutral, the technologies needed must also be exam- ined. To develop these technologies, the challenges for the chemical sciences must be identified. The five most common renewable sources of primary power are hydro- electric, geothermal, wind, biomass, and solar. Hydroelectric power is considered by many to be a model energy source. It is clean, relatively benign environmen- tally, nonpolluting, and relatively inexpensive. However, the global theoretical potential of the hydrology of all the world's precipitation and all of the energy 4Hoffert, M. I. et al., Nature, 395:881. 1998.
36 ENERGY AND TRANSPORTATION flows of the waters on Earth provide only 4.6 TW, far below the goal of 20 TW of power. The technically feasible amount of hydroelectric power is far less, only about 0.7 TW. The installed capacity of hydroelectric power is already 0.5 TW. Therefore, there is not much additional power available for exploitation from this source. Power derived from wind presents significant onshore potential. Two large geographical areas the Great Plains of the United States and the region from Inner Mongolia to northwest China present significant expanses of land suit- able to utilize wind. If 6 percent of the dedicated land in the Great Plains were used for wind farms, approximately 0.5 TW of power could potentially be obtained. As with hydroelectric, this is a significant amount of power but far short of the projected 20 TW of additional power that will be required. Globally, it is theoretically possible to obtain 50 TW of power from wind stations on land, but with practical land usage about 4 percent of all the land that has enough wind to make power generation economically feasible the potential amount of power that can be derived is approximately 2 TW. The off- shore potential for wind power generation is larger, but there is the significant requirement of being close to an electrical grid to make it practical. Distribution is a key concern with wind generation of power. If, for example, the Great Plains were used to generate large amounts of power for the United States, this power would not be consumed locally, and there are constraints to the electrical grid. An efficient method of power storage would have to be found. Presently the methods available for power storage carry too high a penalty in terms of energy loss to make them economically feasible for wind as a significant source of energy. Biomass as a source of a large percentage of the world's power has signifi- cant obstacles. Biomass requires large areas because the process is very ineffi- cient. Only 3 percent of the total sunlight that is incident per unit area on a plant is actually stored in free energy by photosynthesis. This is sufficient for biological needs but is difficult to exploit as a source of primary power. To meet the goal of 20 TW of additional power, biomass would require 4 x 10~3 m2 of land, and the total landmass of the earth is 1.3 x 10~4. Clearly this is not a viable option. It is possible to look at the situation with biomass from a different perspec- tive. The amount of land with crop production potential in 1990 was 2.45 x 10~3 m2. In order to support 9 billion people in 2050, 0.416 x 10~3 m2 of additional land will be required for crop production. The remaining land available for biomass energy would then be 1.28 x 10~3 m2. This would result in a projected total of 7 to 10 TW of power. This would be a massive undertaking, requiring that almost all of the crop production potential on the planet be utilized. In addition, there are significant obstacles, not the least of which is the issue of water resources. Also, cellulose derived from biomass must be readily converted to a liquid fuel preferably ethanol presenting a challenge for the chemical sciences.
R&D CHALLENGES IN THE CHEMICAL SCIENCES 37 Nuclear power, which will be discussed in greater detail in other presenta- tions in this workshop, would require 10,000 new 1 gigawatt nuclear power plants in order to provide 10 terawatts of additional energy. Building one of these power plants every 2 miles along the California coast would provide only 300 of the needed 10,000 power plants. Clearly, it will not be practical for nuclear power to be the sole source of this additional power. Solar power is still another option for noncarbon primary power. Theoreti- cally there is 1.2 x 105 TW of solar energy potential. However, if solar cells are assumed to be 10 percent efficient, realistic land estimates lead to a practical value of 600 TW of available incident solar power, leading to 60 TW of generated power with 10 percent conversion efficiency. To generate 20 TW of power using solar cells with 10 percent efficiency requires approximately 0.16 percent of the world's landmass, including 8.8 per- cent of the landmass of the United States. To generate 12 TW of primary power by this method would require 0.1 percent of the Earth's landmass, including 5.5 percent of the United States. These numbers are still quite large, but compared to other methods for gener- ating noncarbon primary power, solar power appears to be the most compelling method. To achieve anything approaching these numbers using solar energy requires one of three approaches. A low-efficiency low-cost method is through photosynthesis. Alternatively, a highly efficient process yet high-cost method uses photovoltaics. A third method utilizes semiconductor liquid junctions and photo- catalysts with the ultimate goal of using sunlight to split water into hydrogen and oxygen or to make electricity. Both the cost and efficiency of this process are moderate. Production capacity for solar electricity is currently limited to about 100 MW per year, even though it is a subsidized industry. This industry is growing rapidly on the order of 30 percent per year although it must be noted that this is from a small base. Solar electricity currently makes up 0.1 percent of total electricity production. The rate of progress for a variety of different technologies shows increasing photovoltaic efficiency, yet most of these technologies have to contend with physical limitations. Silicon or crystalline semiconductors have very high con- version efficiencies, yet these crystals are very costly to make because the grain sizes must be large. When smaller grain sizes are used, these semiconductors are cheaper to make, but they have much shorter lifetimes. Single-crystal silicon can be replaced with a less expensive organic material, but these organic films cur- rently also have short lifetimes and therefore produce devices with low efficiency. Based on these physical limitations it appears unlikely that, with normal market forces and normal research advances, these technologies will provide the amount of power needed economically by 2050. What is likely to be needed is the implementation of a new solar technology that initially offers less performance
38 ENERGY AND TRANSPORTATION but at a far lower cost, and one that eventually overtakes present technologies as it improves. In terms of technologies for photovoltaics, a grand challenge would be to devise alternatives to massive single crystals, which clearly cannot be produced economically. One way to achieve this goal would be to passivate grain bound- aries in order to get polycrystalline samples to act as a large single crystal. Electron transfer agents could be used to link together grain boundaries, thereby allowing electrons to move from grain to grain without inducing recombination at the grain boundary edges. One particular method for achieving this goal that is currently being investi- gated utilizes titanium dioxide, an inexpensive pigment found in white paint.5 The TiO2 is coated on a glass slide and sensitized with a dye to absorb sunlight. Electricity can be generated with 5 to lo percent efficiency. While there are still questions about the long-term stability of this material, it does represent a new and inexpensive approach to photovoltaics different than technologies presently used. Other approaches with different light absorbers, such as interpenetrating polymer networks, nanocrystals, and other inexpensive approaches to solar energy conversion, should be explored as well. Photoelectrolysis is a technology that converts light into both electrical and chemical energy. Solid SrTiO3 is used in a photochemical cell where it absorbs sunlight and effectively splits water with high quantum yield. Electrolysis of water using this process can be sustained almost indefinitely. However, the band gap for SrTiO3 (3.4 eV), is in the ultraviolet range, and materials with a lower band gap either are not stable in water, cannot sustain the electrolysis of water, and/or cannot absorb sunlight efficiently. Catalysts are also needed to effectively con- vert the photogenerated charge into chemical fuels. SUMMARY To meet the increased demands for primary power in the 21st century, normal economic driving forces appear to indicate that the demand for 28 TW of power without unacceptable environmental consequences could result from a combina- tion of wind, solar, biomass, and nuclear power. However, sources such as wind and solar require new technologies to effectively store and transport power with little loss. Another important consideration with solar power as a source of energy is that it inherently provides electrical power. Only about 10 percent of energy consumption is presently in the form of electricity, whereas the other 90 percent is used for heating, transportation, and industry. Even if electricity were used to meet part of these needs, it would not be used to meet all of the remaining 90 per- 5B. O'Regan and M. Gratzel, Nature, 353:737. 1991.
R&D CHALLENGES IN THE CHEMICAL SCIENCES 39 cent of power consumption unless direct photochemical or efficient electro- chemical methods for energy storage or fuel generation were developed. To make fuels that are storable and transportable, there are two primary chemical transformations to consider. One is the conversion of carbon dioxide to methanol, and the other is splitting water. Methanol could be used in a fuel cell where it is converted to carbon dioxide. This must be converted back to methanol to close the carbon loop. Alternatively, if hydrogen is used as a replacement fuel for carbon-based fuel, a hydrogen fuel cell that utilizes the product of solar or electrical water splitting would be available for transportation. Whichever of these alternatives is ultimately adopted, the need for additional primary energy sources is apparent. In addition, the case can be made for sig- nificant carbon-neutral energy systems in the future. These technologies present significant challenges for the chemical sciences. For solar power, inexpensive conversion systems must be developed that include effective energy storage. Advances in the chemical sciences will also be needed to provide the new chemistry required to support an evolving mix of fuels for primary and secondary energy.