Poseidon Water employees stand between rows of reverse osmosis filters at the Western Hemisphere's largest seawater desalination plant, currently under construction in Carlsbad, California.
Credit:
Mike Blake/Reuters
. Most modern desalination facilities use reverse osmosis, in which water is pumped at high pressure through semipermeable membranes that remove salt and other minerals.
Worldwide about 300 million people get some freshwater from more than 17,000 desalination plants in 150 countries.
In the United States, a $1 billion plant is being built in Carlsbad, California, to provide about seven percent of the drinking water needs for the San Diego region. When it goes online in late 2015 it will be the biggest in North America, with a 50-million-gallon-per-day capacity. And California currently has about 16 desalination plant proposals in the works.
But desalination is expensive. A thousand gallons of freshwater from a desalination plant costs the average US consumer $2.50 to $5, Pankratz says, compared to $2 for conventional freshwater.
It’s also an energy hog: Desalination plants around the world consume more than 200 million kilowatt-hours each day, with energy costs an estimated 55 percent of plants’ total operation and maintenance costs. It takes most reverse osmosis plants about three to 10 kilowatt-hours of energy to produce one cubic meter of freshwater from seawater. Traditional drinking water treatment plants typically use well under 1 kWh per cubic meter.
Membrane Upgrade
Most experts say that reverse osmosis is as efficient as it’s going to get. But some researchers are trying to squeeze more by improving the membranes used to separate salt from water.
Membranes currently used for desalination are mainly thin polyamide films rolled into a hollow tube through which the water wicks. One way to save energy is to increase the diameter of the membranes, which is directly correlated with how much freshwater they can make. Companies are increasingly moving from eight-inch to 16-inch diameter membranes, which have four times the active area.
“You can produce more water while reducing the footprint for the equipment,” says Harold Fravel Jr., executive director of the American Membrane Technology Association, an organization that advances the use of water purification systems.
A lot of membrane research is focused on nanomaterials — materials about 100,000 times smaller than the diameter of a human hair. MIT researchers reported in 2012 that a membrane made of a one-atom-thick sheet of carbon atoms called graphene could work just as well and requires less pressure to pump water through than polyamide, which is about a thousand times thicker. Less pressure means less energy to operate the system, and, therefore, lower energy bills.
Graphene is not only durable and incredibly thin, but, unlike polyamide, it’s not sensitive to water treatment compounds such as chlorine. In 2013, Lockheed Martin patented the Perforene membrane, which is one atom thick with holes small enough to trap salt and other minerals but that allow water to pass.
Another popular nanomaterial solution is carbon nanotubes, says Philip Davies, an Aston University researcher who specializes in energy efficient systems for water treatment. Carbon nanotubes are attractive for the same reasons as graphene — strong, durable material packed in a tiny package — and can absorb more than 400 percent of their weight in salt.
Membranes have to be swapped out, so carbon nanotubes’ durability and high absorption rate could reduce replacement frequency, saving time and money.
Membrane technology all “sounds sexy, but it’s not easy,” Pankratz says. “There are engineering challenges when making something so thin that still maintains integrity.”
Graphene and carbon nanotubes are decades away from widespread use, says Wendell Ela, a University of Arizona chemical and environmental engineering professor. “I do see them having an impact, but it’s a ways out.”
Truby said barriers to commercialization include engineering such small materials and making new membranes compatible with current plants and infrastructure.
“It’ll be key to upgrade systems without tearing [them] down and building a whole new plant,” he says.
(20min delay)
