Costs and operational performance

Recent technical developments and reductions in the costs of all major categories of solar energy technologies have been substantial. First, consider PVs, for which historical and projected costs (in 1990 prices) are shown in Figure 2. In the early 1970s the costs of PV modules were several hundred thousand dollars per peak kilowatt (kWp), and applications were largely confined to aerospace and other specialized uses. By the early 1980s costs had fallen tenfold to around $25,00() to $50.000/ kWp, and by 1990 to $6,000/kWp. and PVs had become commercially viable for a wide range of small scale uses. In the industrial countries, PVs are often used for telecommunications, cathodic protection of oil and gas pipelines, and as a source of electricity in homes and buildings; and in various "luxury" applications. Experiments with PVs as a source of supplementary grid power are also being conducted in several OECD countries with positive results. In developing countries, common applications are for village and domestic lighting, water pumping, battery charging, and supplies to rural health clinics and schools. The effectiveness of applications in developing countries is well illustrated in a recent report by van der Plas, who notes that 20,000 rural households in Kenya have been provided with electricity from PVs in the past five years - more than were newly supplied from the grid. An interesting point about this development was that the PVs were supplied by market vendors at cost (the systems were also taxed), whereas grid-supplied electricity was subsidized. The engineering and economic data suggest that further progress can be expected on at least two fronts:

Figure 1. Land Use by Solar-Thermal and Photovoltaics Versus Land Inundated for Hydropower

· Scale economies and technical progress in production.

World output grew from 1 MW per year 15 years ago to more than 60 MW today, a growth rate greater than 30 percent per year, albeit from a small base. This is still a small market, but the technologies are modular, and the economies of scale and the technical possibilities for batch production have barely been exploited.

· Further developments in cell, module, and systems design, along with improvements in conversion efficiencies. Development of improved materials, use of multi-junction devices and novel cell designs to capture a higher proportion of the solar spectrum, and use of concentrator (Fresnel) lenses to focus the sunlight onto high efficiency cells are further areas of rapid development.

The U.S. Department of Energy has projected that with market expansion, costs should eventually decline to about $2,000 or less per peak kilowatt (including balanceof-systems costs). If this were to happen, which is quite plausible, PVs would become economical for use in grid connected applications in the distribution networks of countries with good solar insulation's; this level of performance would also favour the emergence of independent or "distributed" utilities.

Figure 2. Photovoltaic Module Costs. Actual and Projected. 1970-2015

Progress in solar-thermal schemes has also been noteworthy (Figure 3). They have already been technically proven for large-scale generation, with costs of $3,000 per kW and 12 to 20 US¢/kWh. Steam conditions compare well with those of fossil and nuclear stations, typically 1,000 psi and 700° F. and operational performance is very good. (The availability of the solar fields in the Kramer Junction plants in California is 99 percent.) Costs are still high in comparison with fossilfired power stations, though appreciably lower than the ex post costs of nuclear power plants commissioned in the United States in the 1980s, and they compare favourably with the costs of some hydro schemes in developing countries. Further, as with PVs, scale economies in manufacture and technical possibilities have barely been exploited. For example, the central receiver technologies offer prospects of major efficiency gains and reductions in costs through a significant increase in steam pressures and temperatures. Experience with solar-thermal power stations dates only to the mid-1980s, with only 350 MW having been built.

Research on a range of materials and design concepts is proving fertile, and ample scope remains for further gains in conversion efficiencies, from the present 7 to 15 percent range to the 15 to 30 percent range for both PVs and solar-thermal stations. The potential is especially large in developing countries, where solar insolations are usually high and energy markets are growing rapidly. Significant progress has also been made in secondary sources of solar energy, such as the use of wind and biomass resources for power generation.

The above developments in solar technologies were much stimulated by high oil prices in the period 197385 and attracted the interest of several major companies. The collapse of oil prices in the mid-1980s led some companies to scale back their investment plans, and in some cases to shelve them, but those that continued their programs reduced costs by amounts comparable with the fall in oil prices. Thus, as real oil prices fell by 75 percent between 1980 and 1992 (from $60 to under $20 per barrel), those of PV modules fell by roughly 80 percent. For wind technologies, costs have declined roughly 60 to 70 percent since 1985 (Figure 4), and for solar-thermal by about 50 percent over the same period. Nevertheless, low oil and gas prices make it difficult for solar energy projects to compete commercially with fossil fuels, and presently their main attractions to private investors and users are for small-scale applications, the possibility of a commercial surprise, and their promise as an alternative to fossil fuels should the need arise for environmental or other reasons.

Figure 3. Cost of Electricity from Large-Scale Solar-Thermal Technologies

Several other features related to the costs and performance of solar technologies are worth noting. One is the short lead times, notably for PVs, solar-thermal schemes, and wind power. Construction times for some of the solar-thermal plants in California were as low as 9 months, and PV systems can be installed in yet shorter times. The times typically quoted for wind are similar to those for solar-thermal. The lead times of biomass-fired power generation projects are likely to be longer unless they are based on residues or high-yielding crops (an area of much research interest). Another feature will likely be the comparative ease and speed of decommissioning once a plant has completed its useful operational life. For practical purposes, we are dealing with a "reversible" technology. Finally, solar installations may allow for "live" maintenance (maintenance while the plant is operating) owing to the modularity of the plant; this too should help to improve operational performance and reduce maintenance costs.

Figure 4. Cost of Electricity from Wind Turbines in California, 1985-1995