Inside The Technology That Cuts The Crap From Water.
The membranes that do the hardest work—pulling out viruses, pathogens, and hormone-disrupting chemicals—perform “reverse osmosis,” known in the industry as RO. They simulate the biological filtration process that happens within our cells as fluids flow across semipermeable membranes. Imagine that a hole in the RO membrane is the width of a basketball hoop. On that same scale, a water molecule would be the width of a basketball, but a pathogen or a virus would be the size of a Hummer—it can’t pass through. Nor can pharmaceutical residues, which on this scale can be as large as a Mack truck.
The most difficult thing to remove is salt, because it isn’t suspended in water, but dissolved. That’s why recycling wastewater is about half the cost of desalinating ocean water: Both use RO membranes, but the salinity of ocean water is much higher, so it’s harder and much more energy-intensive to pump it through the tiny holes. The RO membranes are rolled up inside 8-by-40-inch cylinders, thousands of which are stacked and networked together at the Orange County facility. Water is blasted through the modules with 1,000-horsepower pumps.
Before the water gets to the RO stage, it must go through seven other stages of filtration. In one, it enters a “microfiltration” process in which the water is sucked through thousands of tiny, porous straws. In another, it’s zapped with a purifying ultraviolet light. “At each stage of filtration, you’re using methods that remove contaminants with lower and lower molecular weight,” Desai says. “Ultimately, by the time you get to reverse osmosis, nothing gets by but pure water.”
Traditional forms of treating water are much cruder. Instead of forcing it through membranes, most plants in the U.S. today use chemicals. When water is pulled from the Colorado River, for instance, chemicals known as coagulants and flocculants are poured in, causing particles to bind together, separate from the liquid, and settle to the bottom of a holding tank. The water is then removed and disinfected with chlorine.
Another plus of the toilet-to-tap process: Sewage water doesn’t have to be transported long distances. Almost none of the freshwater consumed by the 22 million people of Southern California is local, and the cost of importing it is climbing.
Even though it’s local and cost-effective, toilet-to-tap has been hard to sell to the public. Proposals for major recycled wastewater plants in Tampa Bay, Los Angeles, and Brisbane, Australia, have failed in the past two decades—shut down by public objection to the yuck factor.
A key concern of critics is that wastewater could be more easily contaminated by a pathogen. Everybody in the water industry remembers the largest epidemic of waterborne disease in U.S. history: a 1993 outbreak caused by cryptosporidium, an invisible parasite that made its way into Milwaukee’s public water supply. Almost half a million people were affected with uncontrollable diarrhea; 70 died from dehydration. The disinfectant chemicals used at the water treatment plant weren’t strong enough to kill the parasite.
Because of this concern, the Orange County plant is required to have an “environmental buffer.” After it gets through the RO process, the water that I drank at the end of my plant tour must flow into a local aquifer and mix with that natural supply before it’s pumped into people’s homes. And yet there’s no health reason why processed water can’t be consumed immediately. “No pathogen, including cryptosporidium, can make it past the physical barrier of an RO membrane,” says Mike Markus, general manager of the Orange County Water District. “Nor can pharmaceutical chemicals or endocrine disrupters or virtually all other contaminants.” He explains that as toilet-to-tap technology becomes more widely accepted, it will be cheaper and more efficient to pump directly into people’s homes.
Desai is frowning into his grande black coffee at Starbucks. He’d asked the baristas to leave “room for milk,” but they filled his cup nearly to the brim. “This is one of those inefficiencies that drives me crazy,” he says. “Think of all the millions of people each day who dump out a portion of their coffee to put milk in because it’s filled too high—all the water and heat and fuel and coffee beans that are wasted because of this negligence! Why isn’t there a line printed on every coffee cup that represents a universal standard for ‘room for milk’?” He pauses, and his face brightens. “If you’re interested in efficiency, everywhere you look there are opportunities.”
Desai has the kind of mind that churns out ideas so fast you can almost hear it humming. The son of Indian immigrants, he grew up in Ann Arbor, Mich., where his father was a civil engineer for Bechtel. While a pre-med student at the University of Michigan, he decided he preferred chemistry to biology; he didn’t like laboratories, but he loved business. “I was the guy who was the social chairman of my fraternity, and I worked my way through college as a bartender,” he says. “I wasn’t cut out for a lab coat.”
He got a job at Dow after graduating, selling water treatment products to power and food processing plants. After a few years in the field, he received an MBA from Northwestern’s Kellogg School of Management, where he took night classes. He spent more than a decade at Dow, then got the entrepreneurial itch and left to help build the startup Natureworks, a maker of corn-derived biodegradable plastics. He took annual sales from $10 million to $100 million and then returned to Dow to head the water division.
Dow’s been making water filtration membranes since the 1970s—at first as a means for desalination. By the mid-’80s, it was selling water filtration products to Intel, General Motors, and other companies that wanted reliable streams of hyperpure water in their manufacturing facilities. Today, Dow is the biggest player in the RO membrane market, but it faces stiff competition from Toray Industries and Hydranautics. “They’re like the Toyota, Ford, and GM of membranes,” says Tom Pankratz, a water industry analyst, “but Dow is considered to be out front.”
Desalination is still a huge part of Dow’s water business. Its membranes were used at a $1.5 billion plant that opened this fall in Carlsbad, Calif., which provides almost 10 percent of San Diego’s drinking water, siphoned from the Pacific Ocean. But given the cost advantages of toilet-to-tap, it has far more potential than desalination. Desai expects his sales will eventually tilt heavily toward toilet-to-tap. “When I look at my innovation portfolio now, pretty much everything in it has some connection to recycling wastewater,” he says. “That’s the megatrend.”
The nerve center of the membrane industry is located in an unassuming metal warehouse in Edina, Minn., just outside the Twin Cities. “This is by far the most membrane made under one roof anywhere on the planet,” Desai says. Tens of thousands of RO units are fabricated here every day. The membranes, which are thinner than tracing paper, are layered with mesh spacers that allow the water to flow through; they’re then rolled into cylinders, wrapped in fiberglass, tested, and packaged.
The process is automated. Each RO cylinder is cut, glued, and rolled by robotic arms that look like praying mantises dancing. “The robotics in this facility allow us to scale and build products that don’t fail,” Desai says. “Failure in the purity of the water supply isn’t an option. The membranes have to work. Precision is everything.”
Desai is reluctant to disclose much detail about Dow’s precision manufacturing; there’s a blue line painted on the floor that I can’t cross, and no pictures can be taken. He says Dow lost its patent on RO membranes in the 1990s—a blow to the company but a boon to the larger industry in the long term, driving competition, efficiency, cost reduction, and scale. In that same period, Dow has increased its products’ efficiency. A decade ago, each RO cylinder could filter 4,800 gallons of water per day; now that same-size unit can filter 6,000 gallons.
“Ten years ago there was only one machine in this whole building—the rest was empty,” says Max Fadeyev, a plant engineer. “Now it’s jampacked with equipment running 24 hours a day, seven days a week. To us, it’s seemed like an explosion.” In December, Dow is opening a second large-scale water membrane manufacturing plant, in Saudi Arabia.
For now, Desai is focused on making membrane products for big industrial and municipal water systems, but he predicts the systems will eventually become smaller, serving communities and even individuals. Dow is also investing heavily in decentralized, at-home water recycling for developing-world markets. Bill Gates made a pitch for a similar approach in January 2015 when he blogged about watching a big pile of human feces on a conveyor belt enter a small-scale waste treatment plant built to serve a community. In minutes, the feces was converted into “water as good as any I’ve had out of a bottle,” Gates wrote. “I would happily drink it every day.” He’s funding this “poop water” technology, as he calls it, developed by Janicki Bioenergy in Sedro-Woolley, Wash., for a pilot project in Dakar, Senegal; the self-powered plant treats and boils human waste, distilling the moisture into clean water.
Desai predicts that all water filtration technology may eventually be this accessible. “It’s Moore’s Law,” he says. “What we’re perfecting at a large scale, for big centralized plants, may become affordable and effective enough to use in a decentralized system, household by household, so that we each control and regenerate our own water supplies.”
Importing water to places like Southern California and Texas has historically been cheap, but with climate change, extended droughts, and increasing stress on rivers and lakes, the economics of water are changing. Virtually every city in the world has to start rethinking the foundation of its water supply. “Not every city has an ocean, not everyone has good lakes and rivers,” Desai says. “But everybody’s got sewage.”