Why GE, Coca-Cola, and IBM Are Getting Into The Water Business

Water is becoming a high-stakes business where there’s money to be made everywhere you look — from greasy wool to microchips. | Illustration by Brock Davis

Why GE, Coca-Cola, and IBM Are Getting Into The Water Business

In the rangeland of Australia, sheep get frightfully dirty. They roam the outback among all manner of plants, trees, and scrub; they loll in the dirt; they sleep on the ground; they roll in their own poop. They shower only if it happens to rain.


So when these sheep get sheared — and Australia is still the largest producer of wool in the world — the fresh wool is grubby. Raw wool is called “greasy wool,” because in addition to dirt, the wool is coated with the sheep’s natural protection, lanolin. A specialized industry exists to clean it. The big Michell Wool scouring plant in Salisbury, a suburb north of Adelaide, uses almost a megaliter of water (264,000 gallons) a day — about what 750 families use.

The inside of the Michell Wool scouring factory smells like a farm, a rich odor of sheep, dirt, the outdoors, and wet wool. The machine that cleans the wool, called a scour, is a long line of connected stainless-steel tanks and conveyors that stretches more than half a football field. It is possible to know exactly how dirty the wool is: Weigh it before it’s scoured; weigh it after. On average, the yield is 55% — 100 pounds of greasy wool yields 55 pounds of usable wool and 45 pounds of dirt, debris, poop, and lanolin. Each pound of wool requires 5 gallons of water to get clean, more than twice what your home washing machine uses for a load.

Salisbury gets just 18 inches of rain a year, less than Flagstaff, Arizona, and sits in South Australia, Australia’s driest state. But until just a few years ago, Michell Wool was washing all its wool in the same water Salisburians were using to shower and make coffee — tap water.


“Back in the 1980s, we were using in excess of a gigaliter of mains water a year,” says David Michell, co-managing director of Michell Wool and a fifth-generation member of the founding family. The company, which supplies wool for a range of uses, including Armani couture, had begun to worry about what might happen if wool washing had to compete with residential water use in terms of price, adequate supply, or both. “If there is no water,” says Michell, “there is no business for us.”

Michell and his colleagues were feeling the first tickles of something most of us are utterly unfamiliar with: water insecurity. Just because the big supply pipe from statewide water utility SA Water was coming into the plant and Michell had been buying $1 million (AUD) worth of water a year, that didn’t mean that in a serious drought, the price wouldn’t rise, the supply wouldn’t be sharply limited, or both.

At Michell Wool, the solution to the company’s water anxiety came not from SA Water — “They said, ‘Just keep buying water,’ ” Michell recalls — but from the city of Salisbury. Town leaders were discussing how to dispose of storm-water runoff more effectively, storm water that Salisbury collected in drains and culverts and piped untreated into the ocean 6 miles west. The town started a new kind of water utility, and Michell Wool became its biggest customer. Salisbury started routing some of its storm water into wetlands and reed beds for filtration and created an underground reservoir of reasonably clean water that’s good for all kinds of purposes: watering ball fields, irrigating commercial nurseries, even piping into toilets, and, of course, washing wool.


Salisbury now pumps 2 gigaliters of water (528 million gallons) a year back out in “purple pipes” to customers who can use it instead of mains water from SA Water. (Purple pipes have become the global standard for water that is not potable but clean enough for other routine uses.) Michell Wool alone takes 15% of Salisbury’s purple-pipe water. And it pays two-thirds less per gallon for purple-pipe water than what it paid for tap water from SA Water.

Upon reflection, it is absurd for routinely drought-ravaged Australia to wash wool in drinking water. In fact, almost regardless of resources, it’s crazy to use drinking water for things like watering soccer fields or flushing toilets. It’s just what we’ve gotten used to.

If there is one truly arresting sign that our relationship to water is about to shift in fundamental ways, it comes not from the world of science or climatology, not from United Nations officials or aid workers desperately trying to get water to people in developing countries. It comes from businesses like Michell Wool — and other corporations with water-intensive businesses, such as Coca-Cola — but also those whose water dependence is less obvious, like GE and IBM. They all have that same tickle of anxiety about water security. For business, water management is fast becoming a key strategic tool. Companies are starting to gather the kind of information that lets them measure not just their water use and their water costs but also their water efficiency, their water productivity, how much work they get from a gallon of water, how much revenue, how much profit.


In the past decade, businesses have discovered water as both a startling vulnerability and an untapped opportunity. Monsanto is developing a new line of seeds and crops that require less water. Robert Fraley, Monsanto’s CTO, says, “We believe that by 2030 we can double the yield for many crops, compared to the year 2000.” In the hospitality industry, Celebrity Cruises has replaced ice with chilled river rock for cold food on the main buffet line at breakfast, lunch, and dinner on all nine of its megaships. That saves 2.7 million pounds of ice-making a year for each ship, ice that requires 330,000 gallons of water to be frozen, treated, and then pumped back overboard. In Las Vegas, the folks at MGM Resorts have worked with Delta faucets to prototype new water-saving showerheads. No less a sage than Warren Buffett has quietly realized how the water landscape is changing. In 2009, his company, Berkshire Hathaway, became the largest shareholder in Nalco, a water-services, treatment, and equipment company that has no public profile but 12,000 employees and nearly $4 billion in revenue.

GE Water is an ambitious new division of the global conglomerate, with 8,000 employees at 50 manufacturing facilities worldwide and revenue of about $2.5 billion. GE Water cleans water for a West Virginia coal mine to reuse; GE Water has built the largest desalination plant in Africa, in Algiers; GE Water has created a wastewater-purification plant that produces 172,000 gallons a day of reuse water to keep the fairways and greens lush at Pennant Hills Golf Club in Sydney.

The new business is busy, but it hasn’t grown as fast as GE would like. It turns out that many companies are skeptical about spending money on water when there is no urgent pressure — be it financial, governmental, or scarcity — to do so. “Customers aren’t feeling a cost for their water,” says Jeff Fulgham, chief marketing officer for GE Water, “so they’re reluctant to spend money to improve their situation.”


Every gallon of water we use has an economic value — the value of whatever we can actually do with that water, whether it’s brew our morning coffee, grow an acre of wheat, or make a microchip.

Yet in our homes, our schools, our companies and organizations, we typically behave as if the opposite were true. We act as if clean, on-demand water has zero economic value. Especially in the developed world, the value inherent in water is hidden under a cloak of invisibility. Although the water has indispensable usefulness, it rarely has a price.


What’s often oddly missing from the conversation about the business of water is the price of the water itself. The companies that are taking water seriously today have something at risk — their inability to function without reliable water, or their reputation if they squander or damage local supplies. Some see an opportunity in persuading other businesses to try to understand their water risk.

What is so striking is that businesses that start to take the economic value of water seriously immediately start to use it and think about it differently.

One revealing sign that business has entered a new age of water is water’s sudden appearance in the financial reporting of companies as diverse as Intel and Coca-Cola. Intel’s website now lists the company’s total water use, broken down by each manufacturing plant around the world, including the names of the rivers and aquifers each factory taps. Coca-Cola seems to have just discovered water’s importance. In its 2002 annual filing with the SEC, under the heading “Raw Materials,” the word water does not appear. But in the 10-K filing submitted in February 2010, the “Raw Materials” section begins this way: “Water is a main ingredient in substantially all our products… . our Company recognizes water availability, quality, and sustainability … as one of the key challenges facing our business.”


Coca-Cola, whose reputation has been doubly stung by controversy over its withdrawals of groundwater in India and by a backlash against its surging Dasani and Vitaminwater businesses, has vowed that by 2020, in the words of president and CEO Muhtar Kent, Coke will become “the first major global corporation where we will be water neutral.” Since almost all of Coke’s products end up as pee — Coke’s customers don’t need more than a few hours to close the loop in the water cycle on the soft drinks and water they consume — it’s not quite clear what a “water neutral” Coca-Cola will look like. But the company is gathering, analyzing, and revealing cascades of water data.

Viewed from a certain perspective, Coke’s business is really a water-processing operation. The company needs 333 ounces of water to generate $1 of revenue. Coke says that every liter of beverage it manufactures and sells requires 2.43 liters of water. That represents a 9% improvement over 2004, which translates into 8 billion gallons of water saved a year.

That’s 8 billion gallons Coke didn’t have to buy or pump out of rivers or aquifers, clean to food-manufacturing standards, and then dispose of. Reduce water use 9%, and you reduce a flood of costs. Companies are realizing that the water bill includes the electric, natural-gas, heating-oil, chemical-treatment, and filtration bills. This water focus isn’t trendy green consciousness or corporate altruism, although in the case of Coke, it is vitally important PR. It’s also business.


Coke and Intel aren’t metering their water use with such precision to satisfy their curiosity or to amuse us. They’re doing it because they want to use less water, because they think they may soon have no choice, and because they’ve discovered that simply measuring water use quickly leads to managing it better.

IBM is one of the companies that has discovered something else about water — that measuring and managing water use is becoming a huge business in itself. At the IBM microchip plant in Burlington, Vermont, a factory where the company makes the kind of ultrapure water necessary to produce semiconductors, the staff knows a lot about its water.

For ultrapure water — a liquid so clean it isn’t safe to drink, so clean it requires its own separate factory inside the microchip plant — IBM’s water staff measures 80 characteristics all the time, in real time, including temperature, flow rate, pH level, and clarity. IBM Burlington has wired the plant’s pumps, tanks, and pipes with 5,000 electronic sensors, each of which gathers about one data point a second. That’s a stream of 300,000 data points a minute. (For comparison, the double-deck stock ticker streaming along the bottom of CNBC provides 52 data points a minute.)


Eric Berliner, water and environmental manager at IBM Burlington, is giving a tour of the utilities plant, where water is heated, chilled, pumped, and cleaned to the point at which only microchips can drink it. The plant hums 24 hours a day with the sound of pumps moving water through fat pipes. It has the musty smell that comes from water and metal pipes being in contact for years. Berliner stops in an alleyway deep inside the plant and nods toward the ceiling: an array of six distinct layers of pipe, each crossing over the other, some as big as a person’s waist, some no bigger than a wrist. Many have labels — hot water, chilled water — with arrows pointing in the direction the water is flowing.

“When you start to think like we think,” Berliner says, his eyes tracing the pipes, “you don’t see water in the pipes. You see dollar signs.”


The water bill at IBM Burlington — just to get 3.2 million gallons a day into the plant — is $100,000 a month. The water staff turns plain municipal water into a portfolio of products, depending on whether someone is mixing high-tech chemicals or running air-conditioning chillers. IBM’s utility plant creates nine custom varieties of water. Each brand of water costs 4, 5, or 10 times more than the cost of the raw water itself.

A few years ago, Janette Bombardier, site operations manager in Burlington, and her staff had a revelation: Water is so important that although it seems far removed from the final product, the computer chips, it could actually be a competitive advantage. “We’ve moved from being a facility that makes chips for IBM products to a facility that makes chips directly for the consumer market. We make chips for cell phones, printers, TVs, cameras, and GPS systems. We go head-to-head with other fabricators in the Far East.”

From Bombardier’s perspective, if she and her staff can find ways to use less water, and to make water more smartly, she is directly reducing the cost of IBM’s chips. Wringing expensive water out of the process helps the giant stay nimble.


“All the issues with water, with energy, with the increasing cost to produce water and move water,” says Bombardier, “that’s always inches from my nose.”

The daily water bill at IBM Burlington, including energy and chemicals, is $10,959. Most of the water used each day — 2.2 million gallons — becomes ultrapure water, the most expensive kind. Of the $10,959 bill, $9,300 a day goes to make ultrapure water. That’s the big target for the staff. Not much point in worrying about how much water the toilets use when 85% of each day’s cost is in ultrapure.

That, in fact, is the first lesson from IBM Burlington. It’s not about saving water, per se — it’s about understanding how you use water, where the costs are and reducing those, where the value is and preserving that.


In that sense, IBM Burlington’s water factory is just like the chilled rocks on a Celebrity cruise-ship buffet. You still need the qualities that water provides even as you reduce the amount of water required to produce them. But if Celebrity is working with an inspired idea, IBM Burlington is working from the analytics — the 400 million data points it gathers daily about water, sifting for patterns, trends, and bulges of wasted energy that aren’t being harnessed. That, in fact, is a part of IBM’s business: teaching people to sift huge quantities of data for important insight and then selling them the computers and the software so they can do it themselves.

In the ultrapure-water factory, it’s the mind flip about water that gets you started. You have to take a step back and look at the water cycle as a whole. “One of the most innovative things we’ve done,” says Bombardier, “is take the energy the water inherently has in it, and we use it for other purposes.”

Water comes into IBM Burlington cold from Lake Champlain and the Champlain Water District. It’s so cold that it has to be warmed up before the staff can turn it into ultrapure water. Meanwhile, the factory has 13 massive, two-story-tall chillers using huge quantities of electricity to produce cold water, even in winter.

If it seems stunningly obvious to connect these two problems, well, it’s really not. There’s coldness in the incoming water that for most of its 50 years IBM Burlington wasn’t quite smart enough to use. The coldness is undesirable; IBM spent money getting rid of it. In another part of the 750-acre campus, water had heat in it that was undesirable, and IBM spent money getting rid of it. In most companies, though, there wouldn’t be much of a pipeline connecting the specialty department that creates ultrapure water with the everyday engineering department that is running the air-conditioning systems.

In a plant that already had something like 18 water-plumbing systems, including ones for steam and segregated fire sprinklers, IBM Burlington has created three fresh loops of water to capture cold and heat where they are, and use them where they’re needed. The cold incoming water, for instance, is routed to areas that need chilling. It provides “free” cold, and in the process, it gets warmed up, also for free, so it’s ready to be ultrapurified.

IBM Burlington now also uses the cold air outdoors — abundant in Burlington, where the average high in December, January, and February is never above freezing — to make cold water during the winter, instead of relying on big chillers.

All of this saves water, and it saves all the things water requires to do its job. And the result? Between 2000 and 2009, IBM Burlington cut its water use 29%, saving the factory $740,000 a year in water bills. But here’s where the magic of water really kicks in. Cutting water use by $740,000 also saves $600,000 in chemical and filtration costs each year, plus an additional $2.3 million in electricity and energy costs. For every $1 that IBM Burlington cuts from its basic water bill, it saves $4 more. “We did 50 things to get there,” says Bombardier. “Angles of usage, treatment, energy capture, using less pump capacity, capturing internal pressure that comes with the water in the line — 50 different things.”

As IBM has discovered, the act of measuring alone creates an imperative for curiosity and innovation, for changing behavior. Just as when you keep track of every calorie you eat, you start cutting back. Just as when there’s a real-time miles-per-gallon number on a car’s dashboard, you can’t help but drive in such a way as to keep the number high.

The real inspiration for IBM has been far more dramatic than simply saving water and money. Burlington has helped IBM change the way it thinks about itself. IBM, the computing company, is creating a whole business around water. It wants to do for its customers — for companies, cities, utilities, whole natural ecosystems — what it has done at IBM Burlington.

In most places, in the United States and the rest of the world, water is not smart. Traffic signals have intelligence, as do cell-phone networks, cable-TV networks — heck, even Walmart’s long-haul trucks are connected on an intelligent network. A water network typically moves only water, not any information about it. Even at the simplest level, for instance, most water meters are still read manually, with someone striding along and popping open your water-meter cover.

“Water is not really measured and monitored in a way that allows you to manage it,” says Sharon Nunes, vice president of IBM’s Big Green Innovations effort. “Water is not disappearing. But as it becomes more scarce in more areas, it becomes critical to better manage it.”

What IBM can do is lay down a nervous system of water sensors that feed an array of computers loaded with analytical software, which lets you see and understand your water. This works whether you’re running a microchip factory, as IBM does — or a university or sewage-treatment plant — or trying to understand the hydrodynamics of a whole bay. IBM, in short, wants to usher in the era of “smart water.” In March 2009, the company formally announced the creation of a water-management-services business unit, along with a list of pilot customers and projects. These include a sensor system to monitor Ireland’s Galway Bay, a similar system to model and monitor New York’s Hudson River, and a contract to create an “end-to-end” smart-water utility for the island nation of Malta.

The conventional estimate is that around the world, water is a $400-billion-a-year business. That’s four times the size of IBM’s annual revenue, but that figure includes everything from digging up worn-out water pipes to building billion-dollar desalination plants. IBM says the smart-water market, the information-technology part of water, could be worth between $15 billion and $20 billion a year.

Water consciousness has a kind of infectious quality. In the spring of 2010, Nunes announced a partnership between IBM and a Saudi Arabian research center to develop an inexpensive desalination system that could be powered by solar energy. In the Middle East, of course, where the whole region needs to manage its freshwater with an eyedropper, finding ways to use the sun to make cheap drinking water is a near obsession. What was remarkable about the IBM announcement is that the project relies on combining two unrelated areas of the company’s technology portfolio, microprocessor technology (in a new kind of solar panel) and nanotechnology (in a new kind of desalination filter), in the service of a third: making clean water, a business that IBM wasn’t in just four years ago.

If water is going to get smart, or more pointedly, if we’re going to get smart about water, these are the kinds of cross-disciplinary leaps that are going to be required. Says IBM’s Bombardier: “We are never done. We are never out of ideas.”

Adapted from The Big Thirst: The Secret Life and Turbulent Future of Water, to be published in April by Free Press / Simon & Schuster. © 2011, Charles Fishman.

Watch this week for more from Charles Fishman and The Big Thirst.


About the author

Charles Fishman, an award-winning Fast Company contributor, is the author of One Giant Leap: The Impossible Mission that Flew Us to the Moon. His exclusive 50-part series, 50 Days to the Moon, will appear here between June 1 and July 20.