"This is good stuff," Pete McLaggan says as he downs a cupful of freshly desalinated water from his company's pilot project, a miniature version of the proposed Carlsbad plant. He hands me a cup, too, and I notice a slightly sweet taste. Between the ocean and my cup, the water had to go through a reverse osmosis system, including a series of filters to remove debris, and then 8,000 highly pressurized membranes that remove salt.

Because of such technological improvements, desalination today costs about a third of what it did 30 years ago. Poseidon is so confident about the economics of desalination that it has promised that its customers will never have to pay more for desalinated water than for imported water. As a result, Carlsbad and eight other cities and water districts have signed contracts with the company.

It's a stiff challenge. Colorado River and State Water Project water costs anywhere from $250 to $700 per acre-foot (325,851 gallons). Poseidon, meanwhile, estimates that it can produce water at $946 per acre-foot, and it will get a $250 per acre-foot subsidy from the Southern California Metropolitan Water District (ultimately paid for by water consumers in six counties). That figure is hotly disputed, however, in part because of high energy prices: The Coastal Commission's staff warns that today's desalination cost is closer to $1,400 an acre-foot. (In fact, water from a host of new desalination plants in Australia costs twice that.)

The company predicts that over time, as demand grows, the cost of imported water will ultimately surpass the cost of desalination. But environmentalists and other critics remain dubious, and warn that Poseidon could ultimately cost ratepayers much more than they bargained for. 

Take the Tampa Bay region on the Gulf Coast of Florida, for example. In 1996, the court ordered a reduction in groundwater pumping by more than 30 percent by 2008, because it was drying up wetlands and allowing saltwater to invade freshwater aquifers. Three years later, Tampa Bay Water, the regional utility, opted to build a desalination plant about half the size of the proposed Carlsbad plant. It was to open in 2002 at a production cost of less than one-fourth the cost of the plants coming on line in Asia, the Middle East and the Caribbean. Poseidon was put in charge of the project.

But in the next four years, three of the plant's contractors went bankrupt. Some filters lasted only four days, and membranes that should have lasted five to seven years had to be replaced after two years, recalls Ken Herd, the plant's director from 2002 to 2008. In May 2002, after the second bankruptcy, Tampa Bay Water decided to take control of the nearly half-finished plant. But the problems continued. By the time the plant was ready to operate in December 2007, the water district had sunk another $48 million into it, and the cost of the water had nearly doubled.

Who was to blame? Nobody could agree. But officials in Carlsbad say they'll insulate themselves from the kind of problems that plagued Tampa by giving the company total control. Still, that puts a crucial public resource into private hands, which critics say is dangerous. "As a utility, you can shift the financial risk to the private sector," warns Tampa Bay's Herd, "but you can't shift the responsibility of providing drinking water to your customers to the private sector."

Critics of the Carlsbad proposal, from the Sierra Club to coastal environmental groups like San Diego Coastkeeper and Surfrider, worry about a West Coast repeat of the Tampa Bay fiasco. But they have an even bigger concern: billions of dead fish.

In the early 1970s, researchers discovered millions of dead fish near power plants in the East and the Deep South. The kills were blamed on the practice of pulling power-plant cooling water out of the ocean with large intake pipes, distributing it to the plant through a condenser and discharging it back into the sea at elevated temperatures.

So California fish biologist Pete Raimondi wasn't surprised when, 15 years ago, as he started studying how marine life responded to the state's 22 coastal power plants, he saw oodles of dead fish. Some fish, birds, marine mammals and other large organisms get pinned against or otherwise caught in huge intake pipes. Others squeeze through the intakes, but die inside the facilities, from the heat or from high-speed collisions with one another or by smashing against the sides.

"Everything that goes in, dies," says Raimondi, a University of California at Santa Cruz biology professor who has reviewed Poseidon's plans for the Coastal Commission.

The San Onofre nuclear power plant north of Camp Pendleton kills an estimated 6 billion to 7 billion fish larvae annually, Raimondi says. Another 4 billion a year die at the Carlsbad power plant, as well as at the one in Huntington Beach. Desalination plants can take fish on a similarly fatal ride; in fact, many desal plants share water intake systems with neighboring power plants.

Power plants have tried to mitigate kills by putting screens on intake pipes, and they've tried to offset damage by restoring wetlands elsewhere. Southern California Edison, which operates the San Onofre nuclear plant, is building a 150-acre wetland restoration project in Del Mar north of San Diego. Meanwhile, Poseidon has agreed to construct a 55-acre wetland restoration, but it has yet to specify a site, and critics are dismayed by the fact that it has up to seven years to complete it.