Paul Westbrook is a slender guy who wears his enthusiasm on his sleeve. He’s the kind of guy who articulates every syllable when he talks and can say “goofball” in casual conversation without irony. He plays alto sax in a jazz band with fellow Texas Instruments engineers. He drives a Prius, charting his gas mileage on his personal web site, and he’s the 10-year co-captain of TI’s track and field team. (He’s a sprinter who’s been inducted into the U.S. Corporate Athletics Association Hall of Fame.) He is, in short, an unassuming, likable, multitalented nerd.
He is also, arguably, the reason Texas Instruments’ latest semiconductor plant has garnered the attention and admiration of everyone from New York Times columnist Thomas Friedman to chipmakers in China. The 1.1-million-square-foot wafer plant, also known as a fab, could have been built by cheap labor overseas. Westbrook, a 23-year veteran with the company, was one of many TI engineers to argue for building closer to home. A facility could be cheaper here and run more efficiently if it followed the principles of conservation, recycling, and alternative energy sources, they said. Westbrook had the proof.
“On the engineering level, we realized there was a lot of really good, doable stuff. Then we realized we needed some managerial support,” Westbrook, 45, says. So on a Saturday afternoon in the early fall of 2003, Westbrook gave some higher-ups a tour of his house in Fairview, Texas. He showed them the solar plate used to heat water in the winter and the overhangs used to shade living areas in the summer. He pointed out the ground-source heat pump, the two 1,600-gallon rainwater collection tanks, the argon-filled windows, and dozens of other energy-efficient doodads. Lastly, he showed them his utility bills. In the 10 years he and his wife have lived in their two-story, 2,713-square-foot house, the average monthly electric bill has been $64.
“Do these really scale up?” one of them asked.
“Yes,” Westbrook said. “The principles scale up.”
There was a pause.
“Great. What do you need to make this happen?”
Of course, it wasn’t that simple. Yes, the TI execs were starting to warm to the idea of the new semiconductor plant being built just six miles up Central Expressway from headquarters, provided it made financial sense. Over the years, the company has enjoyed a good relationship with local government officials who are eager to help when they can. Good government relationships make it easier to do business. Plus, the plant workers wouldn’t have to pack up and relocate—not an inconsiderable expense and inconvenience. More importantly, TI research and development would be down the street from production. The only problem? The research and development in Dallas isn’t what it could be.
“Dallas is the ninth-largest city in America, but it lacks a top-50 engineering school,” says TI spokesman Phil Ritter. “In our minds, as we were looking at the site selection, that was a huge deficiency. We asked the state to consider that in terms of their response and their proposal to us to build the plant here.”
The state responded. Governor Rick Perry put together a $300 million package for the nearby UT Dallas engineering school with the intention of making it a Top-50 engineering school in five years. The $300 million will primarily go to research facilities, the recruitment of top faculty, and scholarships for top students. According to projections, the investment will be well worth it. The Perryman Group estimates the economic impact of the plant on the Dallas-Fort Worth area to be 74,000 permanent jobs and $12 billion in spending.
With site selection completed and cost-competitiveness in mind, Westbrook and his fellow engineers set out to design TI’s next semiconductor plant. It would not be easy. Upper management wanted the Richardson plant to cost 30 percent less than the most recent one, built in the mid-’90s.
“We all said, ‘Crazy, goofball management. What are they thinking?’” Westbrook says. But the stringent budget forced the design team to question everything and constantly look for a better, more efficient way to build and operate.
“Cost reduction became a huge friend of sustainable design,” Westbrook says. “It really did.”
No one knows about that friendship more than Amory Lovins. Whereas Westbrook is a cheerleader for sustainability, Lovins is the starting quarterback. The head of Rocky Mountain Institute, Lovins is one of the foremost energy consultants in the world. In 1999, Lovins co-authored Natural Capitalism: Creating the Next Industrial Revolution with his wife, Hunter, and Paul Hawken, the environmentally conscious entrepreneur.
In December 2003, Lovins and about a dozen members of his think tank met with about 30 TI engineers in a windowless room on the ground floor of a TI office building at headquarters for a three-day charrette. After introductions and a short lecture on natural capitalism and “tunneling through the cost barrier,” the group broke up into smaller groups, five or six brains tackling architecture, water process, heating and cooling, and more. Classic rock music played in the background, but the low rumble of communal brainstorming soon overwhelmed it.
By the end of the first day, the room was papered with Post-It Notes offering broad suggestions about water reclamation or the need for natural grass or porous concrete. On day two, the theoretical gave way to the tactical, as formulas and equations accompanied the big picture ideas. Possibility begat creativity and vice versa.
On occasion, the groups toured their neighbors, sharing their likes, concerns, and suggestions. It was mass chaos, but synergy often is. It was also wildly successful. J.D. Bryant, the facilitator of the charrette, has been coordinating group sessions like this one for 10 years. “It was unlike anything I’ve ever seen,” he says. By day three, the group had laid the groundwork for the largest sustainable building in North Texas at the time, putting Texas Instruments in the midst of the next Industrial Revolution.
Big businesses are realizing that resource management isn’t just about hugging trees. It’s about watching the bottom line. When people were relatively scarce and nature was abundant, a business was forced to increase labor productivity to be competitive. Now, nature is relatively scarce and people are abundant, so a business must increase resource productivity to be competitive. “Can you imagine if you can get 10 or 20 times more out of every unit of electricity than we do now? That is a revolutionary kind of concept,” Westbrook says.
The estimated ROI in the first year from efficiency measures at the Richardson plant is $750,000. At full operation, TI expects a savings of more than $4 million a year, thanks to an estimated 20 percent cut in energy use and 35 percent cut in water. Below are but a few examples of how they did it:
>> The best way to rid your building of unwanted energy—like, say, heat in the middle of August—is to prevent it from getting there in the first place. The roof of the Richardson fab is made of a material that reflects 85 percent of the light that hits it. Westbrook estimates that alone saves about 100 tons of cooling. He tested the roof last summer. At 2 o’clock on an August afternoon when the temperature was 95 degrees, the Richardson plant’s roof was 118 degrees. The roof on the plant built in the ’90s—a light-gravel mix—was 140 degrees. A piece of tar paper got up to 152.
>> Heat generation inside a fab is unavoidable, but you can’t let it build unchecked. Worker discomfort is one thing; faulty chip production is quite another. The clean room is a delicate, precise environment. “When this place is up and running, we’ll be exhausting about 600,000 cubic feet a minute that we have to make up with clean air,” Westbrook says. “And we have to keep it at 70 degrees plus or minus a couple of tenths, and 45 percent relative humidity, plus or minus 3 percent.” But rather than exhausting that heat right away, the designers trapped the heat for reuse. Why make a big, expensive boiler do the job of something that’s inadvertently doing it already? Plus, trapping the heat led to less water loss through evaporation. As Westbrook explains the cycle, he remembers another Amoryism: “You know you’re on the right track when you accidentally solve many problems by solving just one.”
WATER WORLD
>> Typical fabs go through a lot of water: about two and a half million gallons a day are used for rinsing, cooling, and heating. Engineer and water-use expert Pablo Ruiz analyzed the water cycles and was able to find secondary and even tertiary uses for water that was previously destined for the “reject stream.” Ruiz and the engineers cut the water usage by about 750,000 to 800,000 gallons a day—about the equivalent of a large water tower. Every day.
>> Much of that savings comes from reclamation. About 80 percent of the land on which the Richardson facility sits drains to the northeast. There the designers dug out a 2.7-million-gallon retention pond, with a capacity for another two million gallons during heavy rains. The water in that pond is then used for irrigation of the entire campus. A submerged air compressor powered by an above-water windmill sends bubbles up through the pond to keep it aerated.
>> The waterless urinals in the men’s rooms attract perhaps the most curiosity and attention. Because the population of fabs is typically 75 percent male, there’s a savings of about 40,000 gallons per urinal per year with the switch to waterless. Plus, they’re not as expensive to install because you don’t need to pipe water through the wall, and they save money on maintenance because there are no flush valves to get stuck.
>> The automated hand-wash sinks are innovative as well, and they happen to be one of Westbrook’s favorite touches. “It’s a little thing but I think it’s so cool and good-thinking,” he says. Instead of typical, electricity-powered sensors or even regular battery-powered ones, the new TI building has tiny water turbines in the stream. “So when you use a sink, it spins the turbine and recharges the batteries for you. It puts about an 8- to 10-year life on your battery. They’re a little more expensive, but I’ll pay for good engineering.”
LET THERE BE (JUST ENOUGH) LIGHT
>> Lighting is the no. 1 cooling load in an office building, a necessary evil that gives off heat as a byproduct to illumination. Most office buildings have two options when it comes to light switches: on and off. At the administration building, the designers invested a little more money to save money. Ergolights have motion sensors to turn themselves off when not in use and can be networked. Each cubicle worker, through his or her computer, can dim their overhead light to a desired level.
>> The fab designers also used natural light to their advantage and mitigated its negative effects. By rotating the footprint of the administration building just 45 degrees, shielding a face of windows from harsh, western sun exposure in the summer, air-conditioning savings amount to about $30,000 per year.
Lovins is impressed. Not as much by the technical, mechanical, and sheer common sense solutions at TI’s new plant. More so by their implementation.
“Paul’s most remarkable achievement was in timing,” he writes in a recent e-mail. “When business is good in the chip industry, companies are frantically trying to build plants to get the next generation chip to market. Fab designers have no time to build anything other than a retread of what’s been built before. When business is bad, there’s no budget for designing, and by the time there is, it’s too late.
“Paul somehow managed to slip through the crack between these two conditions. With astute technical insights and diplomacy, he persuaded his superiors at exactly the right moment that innovation was possible and indeed vital. I’d been trying to do this for a decade with various chipmakers, and he’s the first one to pull it off. Now it’s getting enough attention in the industry that I hope others will follow suit. If they don’t, they’ll lose market share.”
