Salt water and waste heat - presto! - drinking supply

Every day, some 10,350 plants around the world create more than 8.3 billion gallons of drinking water for a growing thirsty population.

They do it by turning salt water into fresh, using steadily cheaper techniques. Now, two engineering professors at the University of Florida have taken that technology a step further with a novel idea.

Since power plants need water for cooling purposes and desalination plants need heat, why not combine the needs of both? The professors - James Klausner and Renwei Mei - calculate that their process would shave a sixth of the cost from today's most efficient technology.

"Water is critical to power production which requires a large amount of it," says Barbara Carney, desalination project manager with the National Energy Technology Laboratory, an arm of the United States Department of Energy in Morgantown, W.Va. Now, "instead of power plants being a net user of water, they will be producers of water."

Currently, desalination plants - most of them located in the Middle East - use one of two processes to turn salt water into fresh. One involves boiling salt water and condensing the vapor to produce fresh water, a process called distillation. The other uses high-pressure pumps to force salt water through fine filters that trap and remove waterborne salts and minerals in a process called reverse osmosis. Both technologies are energy intensive and not cost-effective on a large scale, except in areas such as Saudi Arabia where water is short and energy is cheap.

The new technique - called diffusion-driven desalination or DDD - uses heat wasted by electrical power plants.

Since that heat lacks the intensity to boil salt water, Professors Klausner and Mei simply use it to heat the water. The water is then sprayed into the top of a diffusion tower - a column packed with a matrix that creates a kind of slow-motion waterfall. Meanwhile, warm air is pumped up from the bottom of the tower. As the trickling salt water meets the air, evaporation occurs. The evaporated - and now salt-free - water is captured. "Instead of releasing the evaporated water, it will be condensed to produce fresh water," explains Ms. Carney.

Thermoelectric power plants consume about 39 percent of the water used in the US, second only to agriculture. Most of the water is used for cooling to condense steam. Each kilowatt-hour of electricity requires about 25 gallons of water to produce. So indirectly, Americans may be using as much water when they turn on lights and run appliances as they do when taking showers and watering lawns.

So far, a prototype DDD plant is producing about 500 gallons of fresh water daily. Klausner and Mei calculate that a DDD plant tapping the waste heat from an average 100-megawatt power plant could produce 1.5 million gallons of fresh water daily. The estimated cost: $2.50 per thousand gallons, compared with $10 per thousand gallons for conventional distillation and $3 per thousand gallons for reverse osmosis.

Though DDD plants designed to produce up to 5 million gallons of fresh water daily appear reasonable, "market studies suggest that we have less barriers to market entry when working with smaller facilities," Klausner says. He estimates the cost to build a facility producing 1 million gallons per day would be about $2 million.

Utilities could build DDD plants next to their power stations and take advantage of their waste heat to produce fresh water for sale, he adds. Other industries that produce waste heat and use lots of fresh water - such as refineries, pulp and paper plants, and chemical- and food-processing plants - could also build their own DDD plants and supply themselves.

"We are very interested in moving the technology out of the laboratory into the commercial sector," Klausner says. To do this, the University of Florida is working with Global Water Technologies Inc. (GWT), a water purification company in Golden, Colo., and seeking to license the technology to other firms.

GWT expects to participate in building a large-scale demonstration unit this year, says George Kast, the company's chief executive officer. A larger commercial plant could follow in 2006, he adds.