The SAR notes that climate change would impact energy supply and demand. Subsequent studies confirm this sensitivity. Hydropower generation is the energy source that is most likely to be impacted because it is sensitive to the amount, timing, and geographical pattern of precipitation as well as temperature (rain or snow, timing of melting) (see Chapter 4). Where reduced streamflows occur, they are expected to negatively impact hydropower production; greater streamflows, if they are timed correctly, might help hydroelectric production. In some regions, change of streamflow timing from spring to winter may increase hydropotential more in the winter than it reduces it in the spring and summer, but there is a question of whether the electric system can take advantage of the increase in winter flows and whether storage would be adequate. Hydroelectric projects generally are designed for a specific river flow regime, including a margin of safety. Projected climate changes are expected to change flow regimes-perhaps outside these safety margins in some instances (see Chapter 4). Although it is not yet possible to provide reliable forecasts of shifts in flow regimes for world river systems as a consequences of climate change, what is known suggests more intense rainfall events (which would require more conservative water storage strategies to prevent flood damage), greater probability of drought (less hydroelectric production), and less precipitation falling as snow (less water available during warm months). All three factors point to less (or, at least, less flexible) hydroelectric capacity at current powerhouses. Reduced flows in rivers and higher temperatures reduce the capabilities of thermal electric generation (Herrington et al., 1997); high temperatures also reduce transmission capabilities.
Some advanced energy technologies also may be affected. For example, the United States and Japan are trying to learn how to exploit the potential of methane hydrates. If global warming leads to warmer oceans or warms areas that currently are permafrost regions, these compounds are likely to become less stable, making it more problematic to attempt to recover methane from them (Kripowicz, 1998).
Increased cloudiness can reduce solar energy production. Wind energy production would be reduced if wind speeds increase above or fall below the acceptable operating range of the technology. Changes in growing conditions could affect the production of biomass, as well as prospects for carbon sequestration in soils and forest resources. Climate change could worsen current trends in depletion of biomass energy stocks in Africa, which is expected to become drier (see Chapters 3 and 10). The impact on biomass elsewhere is less clear; it may include enhancement of growth because of higher rainfall in Africa as well.
The portion of total energy supply from renewable energy sources varies among countries, developed and developing. In the United States in 1998, renewable sources provided roughly 7% of gross energy consumption-about half of that as hydroelectric energy (EIA, 1999a). In other countries, developed and developing, the percentages vary. For example, biomass accounts for 5% of north African, 15% of south African, and 86% of sub-Saharan (minus South Africa) energy consumption; in Cote d'Ivoire, the Democratic Republic of Congo, Ethiopia, Mozambique, and Zambia, the vast majority of on-grid electricity generation comes from hydropower (EIA, 1999b). Hydroelectricity represents the primary source of electricity in Canada and most South and Central American countries, with the highest reliance in Paraguay and Brazil (99 and 87% of generating capacity, respectively) (EIA, 1999c). Although renewable energy sources may be adaptable to new climate, larger percentages of renewables (especially hydroelectricity) in a country's energy supply might make the country relatively more sensitive to climate (see Chapters 10, 11, and 12). However, fossil fuel extraction may be adversely affected by increased wind and wave action, heavy precipitation, shoreline erosion, and permafrost melting in regions where this applies (see Chapter 16). In addition, thermal power plants can be adversely affected by loss of cooling water as a result of low flows (see Chapter 12). If a warmer climate is characterized by more extreme weather events such as windstorms, ice storms, floods, tornadoes, and hail, the transmission systems of electric utilities may experience a higher rate of failure, with attendant costs (see, however, Chapter 3 and TAR WGI Chapter 10) These failures can be extremely costly, as illustrated by the great eastern Canada ice storm of January 1998, which toppled hundreds of transmission towers and downed 120,000 km of power lines-in some cases for a month to 6 weeks-and cost CDN$3 billion in economic damage (only half of which was insured) (Kerry et al., 1999). A 5-week power failure in the central business district of Auckland, New Zealand, occurred in February-March 1998 when four high-voltage transmission cables failed (Ministry of Commerce of New Zealand, 1998). Hot weather contributed to high demand and less-than-optimal operating conditions of these cables as a result of high soil temperature and dryness, although it was not ruled the direct cause. Transmission and distribution systems can be hardened to respond to greater risk, but only at substantial cost.
The SAR notes that on the demand side, space-cooling demand would increase and space-heating demand would decrease. Electrical system expansion (generation, transmission, and distribution) may be required to meet greater summer peaks. In warmer areas, it is expected that the demand for electricity will certainly increase, as may the demand for energy overall. Urbanization, rising incomes, and warmer climates could combine to increase energy used for space cooling-already a major concern in tropical and subtropical cities, most of which are in developing countries (e.g., as much as 60% of total electricity use in the commercial sector in Hong Kong, 60% of all electric energy in Riyadh-see Al-Rabghi et al., 1999; Lam, 1999, 2000). Besides having a major impact on the energy sector, air conditioning would tend to enhance heat island effects because of the energy used. At the same time, research has found that air conditioning, where available and affordable, is a statistically significant factor in reducing the chances of hot-weather-related mortality (Chestnut et al., 1998).
The heat island phenomenon may have a positive impact in cities with seasonally cool to cold winters. For example, during the 20th century, the long-term impact of the urban heat island has been to reduce potential energy demand for space heating by as much as 50% in the central quarters of megacities such as Tokyo and Mexico City (Jáuregui, 1998). Urban warming and increased demand for cooling is expected, even though urban aerosol production (e.g., from power plants) does have a cooling effect (Science News, 1992; Jáuregui and Luyando, 1999).
Additional studies that have been published since the SAR continue to show that whether net energy consumption will increase or decrease as a result of climate change depends very much on location-in particular, whether energy consumption includes larger heating loads or cooling loads. The north-south orientation of Japan provides some insight into this question. Ichinose (1996) (quoted in Mimura et al., 1998) has shown for Japan that reduction in heating would be about 30% in Sapporo on the northern island of Hokkaido, whereas it would be only 10% in Tokyo on the central island of Honshu. On the other hand, electricity consumption for cooling would increase hardly at all on Honshu and several percent in Naha on the southern island of Okinawa. The direction of net change also is sensitive to the future market penetration of air conditioning and to energy prices. In one Japanese study (Hattori et al., 1991, summarized in Mimura et al., 1998) the sensitivity of peak electric power demand to air temperature was shown to have increased 2.3 times during the 15 years between 1975 and 1990, largely as a result of the increase in the market penetration and unit size of air conditioners. Amano (1996, summarized in Mimura et al., 1998) points out that a decline in energy prices contributed to the increase in the sensitivity of electricity consumption to climate.
Belzer et al. (1996) is among the few studies since the SAR that has estimated the effect of climate change on energy demand by the commercial sector. The study projects the change in demand at the national level for the United States in 2030. Accounting for changes in the building stock, a 4°C increase in average annual temperature, holding other loads constant, leads to an estimated 0-5% reduction in total energy consumption by the commercial sector in the year 2030 (note that 4°C was then considered possible at mid-latitudes if worldwide temperatures increased 2.5°C; now it is probably at the upper end of potential increases).
The SAR notes that energy used for irrigation would increase. Peart et al. (1995) studied the effects of climate change on energy efficiency in agriculture (including irrigation) in the southeastern United States. Results indicate that climate change would cause an increase in energy inputs required to produce a given amount of maize, soybeans, and peanuts.
Only a handful of studies since the SAR have looked at the effects of climate change on overall energy demand. Mendelsohn and Schlesinger (1999) estimated climate response functions and economic welfare for the entire energy sector in the United States, based on the cross-sectional study of household and firm energy expenditures in Mendelsohn and Neumann (1998). Two approaches were used: laboratory experiments coupled with process-based simulation models, and cross-sectional studies to substitute for impacts of climate change over time. Economic welfare associated with energy was found to have a quadratic relationship with temperature, with a maximum at 10°C. Although the experimental method succeeded in isolating the effect of climate from other variables, it failed to fully incorporate adaptive responses. The cross-sectional studies found that annual energy expenditures were minimized with an annual temperature of 12.8°C in the commercial energy sector and with 11.7°C in the residential energy sector. The cross-sectional approach, of course, does not allow for the transient response of the climate system or the actual dynamics of the energy sector in response to climate. It substitutes static history for a dynamic future and cannot deal with irreversibilities, higher moments of climatic changes such as alterations to diurnal or seasonal cycles, synergic responses (see Section 7.6), or extreme events. (e.g., Schneider, 1997).
Other reports in this collection