The desalinated water supply is expected to grow by 9.5 percent a year and hit 54 billion cubic meters a year by 2020, or triple the amount out there now, according to a new report from Lux Research. You can buy the report, which covers 13 criteria over two axes (which sounds something like 3-D chess). Or, if you're cheap, just read the following excerpts from these articles I wrote:

• Desalination is hot, but still expensive. Desalination expert Energy Recovery pulled off one of the few successful green IPOs in 2008. Over 90 desalination projects have been announced in the last three years.

• The earth pretty much has the same amount of water -- 1.4 billion cubic kilometers -- as it did a few billion years ago. Only about 0.75 percent of that, however, consists of readily accessible groundwater or freshwater, according to the World Water Council. The rest is frozen (2.25 percent) or salty (97 percent.). Thus, if we want more, the sea or brackish wastewater.

• But, ugh, the price. Some estimates put desalination at $650 to $1,000 or more per acre foot when water agencies often sell water for $200 an acre foot. Put another way, a 25 million gallon a day reverse osmosis plant for seawater can cost $100 million. Stink!

• So what do you do? Energy Recovery reduces the cost by reducing the energy required to desalinate. In reverse osmosis, water is pressured to high levels and effectively shoved through a membrane. The company harnesses the pressure from the water coming out of the backside of the membrane and uses it again. It has projects underway all over the world.

• NanoH2O, meanwhile, says it has a better membrane that allows water companies to get by with less pressure. It came out of UCLA. Yale's Oasys Water, by contrast, uses forward osmosis. Salty water is attracted to a highly salty ammonia solution on the other side of the membrane. It is drawn through the membrane through chemical attraction (no additional energy needed). The ammonia salts are then boiled off. In essence, they are purifying water with ammonia -- fancy that.

• Secretive Quos, meanwhile, has graphite membranes. Check out the patents. Khosla Ventures has investments in NanoH2) and Quos. Oasys just closed a $10 million round.

• Then there is the elegant carbon nanotube membrane from Porifera. And let's not forget 212 Resources and Altela, which purify water sludge at petroleum plants.

• But if you're going to invest in desalination, invest in ones (like all of the above) that sell equipment to private firms. If a company needs to rely on deals with municipal water deals, it will be stuck in a Sisyphean cycle of beta tests.