Journal Issue

Where We Stand with Renewable Energy

International Monetary Fund. External Relations Dept.
Published Date:
January 1993
  • ShareShare
Show Summary Details


Each year, the earth receives an energy input from the sun equal to 15,000 times the world’s commercial energy consumption and 100 times the world’s proven coal, gas, and oil reserves. Modern solar electric schemes are capable of converting 10-20 percent of the incident energy into a form useful for consumption, and in theory they would need less than 1 percent of the world’s land area to meet all its energy needs—which is less than the land areas now occupied by hydro reservoirs, and not much more than is planted for potatoes. Yet, despite the abundance and attraction of the solar energy resource, only a tiny fraction is used.

The situation, however, is changing. The last two decades have seen major technological developments for harnessing solar energy in ways that have greatly increased conversion efficiencies and reduced costs. Several of these developments were not predicted: for instance, the discovery in 1974 that an approximately 9:1 alloy of silicon and hydrogen is a semiconductor, or the possibility that the costs of photovoltaics (PVs) could fall by a factor of 50-100 in the course of two decades.

A once exotic and highly expensive source of electric power, suited mainly for satellites in space, is now an increasing source of electricity in villages, industry, and homes, and prospectively for large-scale power generation. When oil prices reached $30 a barrel, there was considerable commercial interest in renewables in the energy industry. Although this interest abated somewhat when oil prices later collapsed, the abatement was not great, as the costs of renewables have since declined too, with technical progress.

Renewable energy sources range from solar energy to various methods of harnessing wind, geothermal, and wave energy. Already, there are many commercial applications of each of these, but in this article, we will just focus on three technologies: solar-thermal and PV schemes for electricity generation, and biomass (crops and woody material) for the production of electricity and liquid fuels.

Why renewables?

Until recently, the argument for alternatives to fossil fuels—the justification for the nuclear power programs of the 1950s, 1960s, and 1970s—was that a backstop technology was needed in case we ran out of them. But the world’s proven reserves of fossil fuels are very large, over 800 billion tons of oil equivalent energy (t.o.e.), of which 70 percent is coal and 30 percent oil and gas—enough to last a century at today’s levels of consumption and for 50 years allowing for the growth of demands in developing countries. Further, these are only the commercially proven reserves, which have expanded appreciably for many decades. Industry estimates of ultimately recoverable reserves are about 4,600 billion tons, including 1,400 billion tons in oil shales and tar sands—sufficient to last us for the next 150 years or so, assuming continued growth in world demand over the next century. Thus the old backstop argument is no longer valid. Rather, the case is being made on three other grounds.

Economics. The economic case is that the technologies will eventually compete with fossil and nuclear fuels—and also with hydro-electricity. Indeed, they are already competitive for smaller-scale applications, and markets are growing. As the World Bank’s World Development Report 1992 noted, world energy demands are likely to double in the next 30 years and then treble to 20 billion t.o.e. in the next 40 years, even under an “energy efficient” scenario. This will put some pressures on costs and prices as the lower cost reserves are exploited and markets turn increasingly to synthetic fossil fuels in the oil and gas industries. It is worth remembering that a world energy market of 20 billion t.o.e. translates into a market of over $4.5 trillion per year at current world prices, more than half of which will be in the developing countries, and the economic gains from new, lower-cost energy resources would be considerable. Solar power would be especially suited to developing countries: there are huge areas where the incident energy (insolation) is 2,000-2,500 kilowatt-hours (kWh) per square meter per year—twice the levels found in the United Kingdom, the Netherlands, and Germany, for example.

Environment. That solar schemes have no net emissions of carbon dioxide, particulate matter, sulphur dioxide, or nitrous oxide is a commonly cited advantage, and there is little doubt that private investment and research and development (R&D) would be stimulated by the introduction of environmental taxes and regulations. In the case of carbon emissions, renewable energy is the only alternative currently available for development, other than nuclear energy from fission and fusion, to stabilize carbon accumulations in the atmosphere should the need arise. Using energy more efficiently will help reduce the rate of growth of energy demands, but emissions are likely to grow even in an energy efficient scenario, and greater efficiency by itself will not solve the carbon accumulations problem. The evidence is still awaited as to the likely extent and consequences of global warming, but investments in renewables must be a required element in any precautionary policy. For example, a 25 percent level of savings from energy efficiency, equal to the whole of the world’s energy consumption today, would still leave the world’s energy consumption equal to three times today’s level in 40 years.

Land andpeople. It is necessary to dispel the commonly held notion that solar energy is too diffuse to harness and would require too much land. For solar-thermal and PV schemes, land requirements are comparatively small. While they are greater than those of coal-fired plants, for example (excluding the mining area), they are considerably less than those of hydroelectric schemes. They typically require only one fiftieth to one twentieth of the land needed for hydro schemes—for the Aswan High Dam in Egypt. only one hun-dreth (see map). There is also much flexibility in the choice of site: the technology is modular, and the schemes can be located in arid areas with low population densities and need not compete with agriculture or forests or people for land. They would also yield very high levels of energy per hectare (e.g., the yield of a solar farm would be over 500 t.o.e. per hectare per year, assuming 10 percent conversion efficiencies, about 100 times that of biomass energy crops).

Solar-thermal energy

The idea behind solar-thermal electric schemes is an old one. In fact, a 45 kWh steam plant was operating in Egypt in 1912, but was shut down during World War I because of cheaper fuel prices. The sun’s energy is concentrated by mirrors onto a receiver that contains a fluid. The heated fluid is then used to raise steam to drive a turbo generator; the fluid may also be a gas that operates an engine directly. There are three main types of concentrator systems: the parabolic trough, the parabolic dish, and the central receiver (see drawing on next page). the system most used so far is the parabolic trough; 3S1 megawatts (MW) of capacity are to be found in California, supplying electricity to the grid. Central receiver technologies are also promising for large-scale power plants; as they can raise steam to high temperatures, they offer prospects of good conversion efficiencies.

Solar-thermal schemes have another attractive feature, of much importance in the long-term, which is that their heat energy can be stored (e.g.. in rocks, oil, sand, salts, or even water). Hence it is possible to use them in the evenings or when the sun is obscured.

Costs are still high relative to conventional schemes, around 10 to 20 US cents per kWh, or about twice the costs of fossil plant. However, the technology is modular and well suited to mass production, and since less than 400 MW have been produced so far, scale economies in manufacturing have yet to be obtained. In this light, cost projections of the US Department of Energy and major energy research laboratories in the United States and Europe, of around 5 cents per kWh, are not implausible—this is lower than the costs of nuclear generation and comparable to the costs of a high efficiency fossil-fired plant.

Photovoltaic energy

Developments in PVs in the past two decades have been profound, as unit costs have declined by almost two orders of magnitude in a 20-year period (see chart). This is roughly the ratio of the average speed of a jet aircraft today to that of a stagecoach (including stops) 150 years ago—an immense technical accomplishment, especially considering that public policies have actually worked against renewables by heavily subsidizing fossil fuels and nuclear energy.

How much land would a solar scheme need?

That solar energy is not a land intensive resource can be seen by comparing its land requirements with those of hydroelectric schemes. The square shows the area that a solar scheme of 4.080 peak megawatts, generating the same amount of electricity as the Aswan scheme, would occupy. (A net conversion efficiency of 10 percent and annual insolation of 2,500 gigawatt-hours per square kilometer are assumed.) The installed capacity of the Aswan scheme is 2,715 peak megawatts, and it generated 10,200 gigawatt-hours last year.

The photoelectric effect was first described in 1839 by the French physicist Edmond Becquerel. In its modern application, in PV cells, light shining on a semiconductor material frees electrons from within the material from fixed sites; the cells are so designed that the electrons cannot easily return to these sites except by flowing through an external circuit, thus generating a current. The cells are connected to each other, packaged in a protective seal, and sold as “modules,” which can be further aggregated into panels and arrays for use in higher-voltage applications.

In the early 1970s, costs were over $300,000 per kilowatt peak (kWp) for the module only, but since then they have dropped to around $6,000 kWh (including balance of systems costs such as structures and dc/ac converters, they are closer to $10,000/kWp). This is still too high for large-scale power generation. However, there has been a rapid growth of markets—from 1 MW in 1978 to over 60 MW in 1992—for remote and “off-grid” applications, such as telecommunications, health clinics, village lighting, and solar water pumps for irrigation and village water supplies. In the United States, PVs are also being used increasingly for providing supplementary power on distribution networks. These markets are still small in relation to the size of the electricity supply industry, about one thousandth of the annual demands for new generating capacity. But over 40 major manufacturers are investing in the technology, and several are gearing up for a second stage of larger-scale production, in anticipation of larger markets and further cost reductions.

A striking feature of the PV market is the wide range of approaches being explored—not only are several materials being tested and used, but alternative qualities of the same material are being tried. The competition for ideas is intense (always a healthy sign), and it is impossible to determine which ones will eventually dominate. First generation PV cells were based on thick-film single crystal silicon cells with high conversion efficiencies, the power source for most satellites. Since the mid-1980s, however, thin-film, amorphous silicon cells have come to occupy one third of the market; they are less costly, but have lower efficiencies. Further developments include:

  • the use of lenses to concentrate sunlight onto high-grade, high-efficiency cells (so-called concentrator systems);

  • methods to increase the probability of photon capture by changes in the cell’s surface shape;

  • the use of multijunction devices to capture a greater portion of the solar spectrum; in the case of amorphous silicon cells, this has doubled efficiencies to 10 percent; and

  • innovations in manufacturing to introduce batch production—something that is more likely as markets expand (scale economies have hardly yet been exploited).

The current efficiency of conversion of light to electrical energy for modules in the field is 3-17 percent, out of a maximum theoretical efficiency for high-grade cells of 47 percent. This is lower than the efficiency being reached in laboratories, which is 6-34 percent (depending on cell type). The lag between laboratory and field cell efficiencies is around seven years.

A long-standing problem with PV electricity is the cost of storage, particularly in off-grid applications. The common lead acid battery is still the option most used, but this is impractical, of course, for large-scale power generation. Much will depend on the development of the fuel-cell—a major area of research now for industrial countries interested in marketing an electrical car.

Biomass energy

Unlike PV and solar-thermal schemes, which capture the sun directly, biomass converts the carbon dioxide in the atmosphere into sugars and then re-releases the carbon dioxide and energy when burned. It can be used directly to produce electricity—as in the numerous cogener-ation plants found in agricultural regions, using the wastes from agro-industries—or it can be transformed to produce alcohol and gaseous fuels.

On the electricity side, a promising new development is the use of biomass gasification methods in electric power plant cycles to take advantage of the high efficiencies of combined cycle technologies. Emission control technologies are available to eliminate particulates and reduce nitrous oxides (there is virtually no sulphur). And if the biomass is produced in a sustainable way, the net emissions of carbon dioxide are negative on account of the enlarged standing stocks of biomass (the greater the number of trees, the more carbon that is stored). On the liquid fuel side, Brazil is now the biggest user of liquid fuel from biomass, in this case, from sugar cane.

Traditionally, there have been two key problems with using biomass as an alternative energy source. First, its cost was high relative to gasoline, although it is not widely enough appreciated how much costs have declined in real terms since the 1970s, owing to technical developments in fermentation and other parts of the process. In fact, costs were beginning to compare well with gasoline until the collapse of oil prices in 1986. Nevertheless, it will be difficult for ethanol to compete in the liquid fuel markets, and the best opportunity for biomass is probably in the higher value-added markets for electricity generation.

Second, it is land intensive. The maximum theoretical conversion efficiency via photosynthesis to useful biomass energy is 6.7 percent for crops such as maize and sugarcane and 3.3 percent for rice, wheat, and trees. However, in practice, the rate is only 0.2-3.0 percent. Thus, it is often rightly argued that the technology is best used, when it is to be used at all, in areas where the crops can serve other purposes, too (e.g., with tree plantations to restore degraded watersheds). This would also increase the economic returns to the projects. Further, a mixture of species is recommended, even if this raises costs, to avoid the environmental dangers of the monoculture syndrome.

Energy policies

What can be done to encourage the development and wider use of renewables in a way consistent with the aims of good policymaking? First, the industrial countries in particular need to diversify their R&D portfolios. Not only does solar energy receive funding that is minuscule compared with fossil and nuclear technologies (about 5 percent of public R&D in energy) but its share of a declining budget has also been shrinking for the past 13 years. International collaboration on R&D also needs to be promoted, as it is in other areas such as agriculture.

Photovoltaic module costs

Source: Based on a review ol over 50 studies and manufacturers by Kulsum Ahmed {lonhcoming).

Note: All costs lor years up 10 and including 1992 are actual: Ihoso after 1992 are projected. The spread* in the points reflects the spread in costs of diflereni technologies, which are al different stages of development. The size of the module used also affects cost, as does Hie size of the order.

Second, renewables would undoubtedly benefit from the adoption of commercial pricing policies for energy, along with an end to the deformities in the level and structure of energy prices arising from government controls and public monopolies. The costs of supplying peak-load electricity is 15-20 cents per kWh in many countries, not far from the costs of solar-thermal schemes with short-term storage. But few countries have adopted a policy of peak-load pricing and thus provide little incentive for developing storage systems. For large-scale power generation, where solar-thermal has immediate promise, and PVs medium-term

promise, applications in developing countries are being stifled by the widespread subsidy of generation by hydro and fossil fuels. In many countries, prices average less than half of supply costs—around 4 cents per kWh when costs are typically over 10 cents per kWh.

At the international level, the Global Environment Facility (GEF) is providing an excellent opportunity to help test and develop renewable technologies in the developing countries. The facility is run jointly by the World Bank, the UN Development Programme (UNDP), and the UN Environment Programme, and was set up to finance the incremental costs on projects with global environmental benefits (in the areas of global warming, the destruction of biodiversity, ozone depletion, and the pollution of international waters). Already it is funding several renewable energy projects, such as the gasification of wood chips and sugarcane bagasse for power generation in modern gas turbines in Brazil, PV schemes in India and Zimbabwe, and several other technologies, such as wind (in Costa Rica) and geothermal energy. The Bank, UNDP, and several development agencies also are financing small-scale applications of renewable energy in many countries.

By facilitating market applications, the GEF and development finance will help to reduce costs and demonstrate the technologies, and, as noted, these technologies may well become justified on economic grounds alone. If so, this will not be the first example of a policy intended to address an environmental problem containing the seeds of an economic surprise.

For more information, seeRenewable Energy, Sources for Fuels and Electricity, edited by Thomas B. Johansson, Henry Kelly, Amulya K.N. Reddy, Robert H. Williams and Laurie Burnham, Island Press. Fort Myers Beach, FL, 1993.

Other Resources Citing This Publication