Water Sustainability in the United States and Cooling Water Requirements for Power Generation

n a national scale, U.S. water use datacollected at five-year intervals by theUSGS from 1950 to the present shows thatsince about 1980 freshwater withdrawals haveleveled off, even as population and gross domesticproduct have continued to increase (Gleick 2002).However, national data are of limited use in ananalysis of long-term water sustainability becausethe aggregation of data can mask changes in demandand potential scarcities that occur at more localizedscales. This factor is especially important in theUnited States because of the climatic variability andthe differing rates of growth in various regions. Toaddress the issue of long-term water sustainabilityacross different geographic regions, we conducteda national-scale study at the greatest resolutionpossible given available data and with a special focuson identifying areas likely to have limited wateravailability as well as increased electricity demands.Data pertaining to water use were collected andorganized at the same spatial resolution, that ofcounties across the continental United States (3,114counties in the lower 48 states). The USGS was theprimary source of the water withdrawal data (USGS2002). These were supplemented by data on climatefrom the Climate Prediction Center of the NationalOceanic and Atmospheric Administration (NOAA),on population from the US Census Bureau, onelectricity generation from the Department ofEnergy, and on agricultural activity and land use fromthe US Department of Agriculture. Using these data,we developed several metrics to characterize wateruse, including the volume of water withdrawn in acounty compared to the available precipitation(defined as the difference between precipitation andpotential evapotranspiration in months where the termis greater than zero), the percent of water withdrawnby various sectors of the economy, the contributionof groundwater withdrawal, the stored-waterrequirements for the driest months of the year, andthe rates of water withdrawal for domestic use,thermoelectric cooling, and irrigation. As an example,a map of the total freshwater withdrawal fromsurfacewater and groundwater sources as apercentage of available precipitation is shown inFigure 1. Areas where this ratio is greater than 100(i.e., where more water is used than is locallyrenewed through precipitation) are indicative ofbasins using other water sources transported bynatural rivers or man-made flow structures. In somecases, a ratio greater than 100 may also indicateunsustainable groundwater withdrawal. Areas wherethis ratio is high are concentrated in the westernUnited States, most notably in the southwesternregions. Maps of other metrics are presented inEPRI (2003).Projected water withdrawals for the year 2025were calculated using current data and assuming a“business-as-usual” scenario, where the rates ofwater use for per capita domestic use and powergeneration per megawatt-hour remain at their 1995values. Total withdrawals for other sectors of theeconomy (agricultural, commercial, and industrial)are assumed to remain at their current levels, broadlyin line with available data for the past two decades.However, the expected increases in population andpower generation lead to substantially increasedwater withdrawals for these sectors. To evaluatedomestic water demand in 2025, we estimate thatthe population in each county will exhibit the same