To Build a (Better) Fire

A kind of hippie Manhattan Project in rural Oregon tackles climate change, air pollution, and deforestation by bringing together the best minds in the field to invent cheap, durable, clean-burning stoves for 3 billion people.

By Burkhard Bilger

Two men walked into a bar called the Axe and Fiddle. It was a Thursday night in early August in the town of Cottage Grove, Oregon, and the house was full. The men ordered drinks and a vegetarian Reuben and made their way to the only seats left, near a small stage at the back. The taller of the two, Dale Andreatta, had clear blue eyes and a long, columnar head crowned with gray hair. He was wearing a pleated kilt, festooned with pockets and loops for power tools, and spoke in a loud, unmodulated voice—like a clever robot. His friend, Peter Scott, was thinner and more disheveled, with a vaguely Biblical look. He had long brown hair and sandaled feet, sun-baked skin and piercing eyes.

The featured act at the bar that night was a burlesque troupe from New York called Nice Jewish Girls Gone Bad. Just how they’d landed in the Oregon woods wasn’t clear, but they stuck stubbornly to their set list and met with only polite applause. Finally, near the end of the show, one of the performers—a spindly comedian with thick, black glasses and a T-shirt that said “Freak”—peered out from under the spotlight and fixed her eyes, a little desperately, on Peter Scott. “Do you have a job?” she said, almost to herself.

Scott said no, then yes.

“That sounds fishy. What is it you do?”

Scott fidgeted for a second, then mumbled, “I make stoves for Africa.”

“You what?”

“I make stoves for Africa.”

Scott was being modest. In the small-but-fanatical world of stovemakers, he is something of a celebrity. (“Peter is our rock star,” another stovemaker told me.) For the past seven years, under the auspices of the German technical-aid agency GTZ (now GIZ), Scott has designed or built some 400,000 stoves in 13 African countries. He has made them out of mud, brick, sheet metal, clay, ceramic, and discarded oil drums. He has made them in villages without electricity or liquid fuel, where meals are still cooked over open fires, and where burns are among the most common injuries and smoke is the sixth-leading cause of death. In the places where Scott works, a good stove can save your life.

He and Andreatta were in Cottage Grove for Stove Camp. A mile or two from the Axe and Fiddle, a few dozen engineers, anthropologists, inventors, foreign-aid workers, and rogue academics had set up tents in a meadow along a willowy bend in a fork of the Willamette River. They spent their days designing and testing wood-burning stoves, their nights cooking under the stars and debating thermodynamics. Stove Camp was a week-long event hosted by the Aprovecho Research Center—the engineering offshoot of a local institute, education center, and environmental collective. Now in its tenth year, the camp had become a kind of hippie Manhattan Project. It brought together the best minds in the field to solve a single, intractable problem: how do you build cheap, durable, clean-burning stoves for 3 billion people?

A map of the world’s poor is easy to make: just follow the smoke. About half the world’s population cooks with gas, kerosene, or electricity while the other half burns wood, coal, dung, or other solid fuels. To the first group, a roaring hearth has become a luxury—a thing for camping trips and holiday parties. To the second group, it’s a necessity. To the first group, a kitchen is an arsenal of specialized appliances. To the second, it’s just a place to build a fire.

Clean air, according to the U.S. Environmental Protection agency (EPA), contains fewer than 15 micrograms of fine particles per cubic meter. Five times that amount will set off a smoke alarm. Three hundred times as much—roughly what an open fire produces—will slowly kill you. A well-made stove can easily clear the air by piping the smoke out through a chimney or burning the fuel more efficiently. Yet most appliance manufacturers see no profit in making products for people who can’t pay for them. And most aid agencies have found easier ways to help the poor—by administering vaccines, for instance. Stovemakers are a chronically underfunded bunch, used to toiling in the dusty margins of international development. Aside from a few national programs in Asia and the Americas, their projects have tended to be small and scattershot, funded a few thousand stoves at a time by volunteers and NGOs. “We’ve been watering this rock for a long time,” Dean Still, the head of Aprovecho, told me.

Lately, though, the rules have changed. As global temperatures have risen, the smoke from Third World kitchens has been upgraded from a local to a universal threat. The average cooking fire produces about as much carbon dioxide as a car, and a great deal more soot, or black carbon—a substance 700 times as warming. Black carbon absorbs sunlight. A single gram warms the atmosphere as much as a 1500-watt space heater running for a week. Given that cooking fires each release 1,000 to 2,000 grams of soot in a year and that 3 billion people rely on the fires, cleaning up those emissions may be the fastest, cheapest way to cool the planet.

DeanStill 282x300 To Build a (Better) Fire

Dean Still. Photo courtesy of Aprovecho Research Center

The engineers of Stove Camp, in other words, found themselves suddenly, blinkingly, in the spotlight—like raccoons caught digging through a scrap heap. “Kill a million and a half people, and nobody gives a damn,” one government official told me. “But become part of this big climate thing, and everyone comes knocking at your door.”

On the first morning of camp, the bleary-eyed participants gathered around some mismatched Formica tables in the lecture hall—formerly a meat locker in a slaughterhouse. Dean Still stood before them in a baggy T-shirt and shorts, wielding a pink Magic Marker. He had the look and manner of an old hippie, with his mop of white hair and wide, walrus mustache; yet he had trained a generation of stove designers and built one of the world’s premier stove-testing facilities. “This is now the definition of a good stove,” he said, turning to the whiteboard behind him and scrawling out some bullet points:

1. Reduces fuel use by more than 50 percent.

2. Reduces black carbon by more than 60 percent.

3. Reduces childhood pneumonia by more than 30 percent.

4. Affordable (US$10 retail or less).

5. Cooks love it.

6. Gets funded.

No stove had ever met all six criteria at once. “So this is what we have to do this week, my dears,” he said. “Save the damned world.” He grinned. “I mean, you didn’t want an easy problem, did you?”

Building a stove is simple. Building a good stove is hard. Building a good, cheap stove can drive an engineer crazy. The devices at Aprovecho looked straightforward enough. Most were about the size and shape of a stockpot, with a cylindrical combustion chamber and a cooking grate on top. You stuck some twigs into the chamber, set them on fire, and put your pot on the grate—nothing to it. Yet one stove used a pound of wood to boil a gallon of water, and another used two. Fire is a fickle, nonlinear thing and seems to be affected by every millimeter of a stove’s design—the size of the opening, the shape and material of the chamber, the thickness of the grate—each variable amplifying the next and being amplified in turn, in a complex series of feedback loops. “You’ve heard of the butterfly effect?” one engineer asked me. “Well, these stoves are full of butterflies.”

Like science and religion, stove design is riven into sects and disciplines. Some engineers use only low-cost materials such as mud or brick; others dabble in thermoelectric generators and built-in fans—cleaner and more efficient, but also more expensive. Most stoves are built for combustion: they consume the wood, reducing it to ash. But a few are designed for gasification instead. These stoves heat the wood until it releases its volatile compounds, which are ignited in the air. (All that’s left of the wood afterward is its carbon skeleton, which can be burned separately as charcoal or used as a fertilizer.) Gasifiers can be remarkably clean-burning, but they’re also finicky. Because the fire burns at the top of the stove rather than rising up from a bed of coals at the bottom, its flames are easily stifled when new fuel is added, turning the stove into a smoke bomb.

In the vestibule of the Aprovecho building, Still had set up a small “Museum of Stoves” on facing wall racks. Its contents came from more than a dozen countries, in an odd menagerie of shapes and sizes: an elegant clay chulha from India, a squat steel Jiko from Kenya, a painted coal burner from China (like an Easy-Bake oven). Most were better than an open fire, yet all had failed the test in some way—too flimsy or inefficient or expensive or unstable or unclean or hard to use. “We still haven’t cracked the nut,” Peter Scott said.

Scott had come to Stove Camp to build a better injera stove. Injera is the spongy pancake that Ethiopians eat with almost every meal. The batter, usually made of an ancient grain called teff, is fermented until it’s bubbly and tart. It’s poured onto a ceramic griddle, or mitad, then set over an open fire or a concrete hearth. In Ethiopia, injera is often cooked by women’s cooperatives in kitchens that may have 40 or 50 smoky, inefficient stoves running simultaneously—one reason that the country has lost more than 90 percent of its forests since the early 1960s. “In the north, people will travel hundreds of kilometers to get wood, then double back to bring it to market,” Scott told me. A good stove, he figured, could cut that fuel use in half.

For the past several months, Scott and his kilt-wearing friend, Dale Andreatta—a mechanical engineer from Columbus, Ohio, who often did stove projects pro bono—had been collaborating on a prototype. It had an efficient ceramic combustion chamber shaped like a miniature fireplace, with a round griddle perched above it like a tabletop. Scott had tried using a traditional mitad, since local cooks would much prefer it, but the ceramic wouldn’t heat evenly—so he’d switched to steel instead. Steel conducts heat much more efficiently than ceramic, and it’s often used for the plancha griddles in tortilla stoves. Injera, though, is an unforgiving dish. Its batter is thin and watery, so it can’t be moved around like a tortilla; any hot spots in the griddle will burn it. “The Ethiopians are unbelievably particular,” Scott said. “If the injera doesn’t have the exact size of bubble in the batter, they’ll say it’s garbage.”

use this article in the classroom To Build a (Better) FireOver the next few days, I’d periodically find Scott and Andreatta skulking around the Aprovecho workshops and laboratories, looking for tools or discussing metallurgy. Their preliminary tests had not been encouraging: the griddle was 200 degrees Fahrenheit hotter at the center than at the edge. When I asked Andreatta how it was going, he lifted an eyebrow. “The optimist thinks the glass is half full,” he said. “The pessimist thinks the glass is half empty. The engineer knows the real truth: that the glass is twice as large as it should be for optimum utilization of resources.”

The trouble with tradition, stovemakers have found, is that it can be remarkably thick-headed. Ignore it, and your shiny new stove may get turned into a flower pot. Cater to it, and you may end up with a new version of the same old problem. The campers in Cottage Grove spent half their time agonizing over cultural sensitivity. “We’re highly dominated by elderly, white, engineering types,” a stovemaker who’d worked in Uganda told me. “So you get a lot of preposterous ideas that’ll never fly in the kitchen.” The other half groused about “design drift.” Too many stoves start out as marvels of efficiency, they said, and are gradually modified into obsolescence. Once the engineer is gone, the local builder may widen the stove’s mouth so it can burn larger sticks, only to draw in too much cold air. Or he’ll make the stove out of denser bricks, not realizing that the air pockets in the clay are its best insulation. The better the stove, the tighter its tolerances, the easier it is to ruin.

“When we first got into this, we had this utopian vision of working with local communities to build locally grown stoves,” Jacob Moss, a stove camper who works for the EPA and started its Partnership for Clean Indoor Air, told me. “We’ve moved away from that—I won’t say 180 degrees, but maybe 160. I don’t really listen to small-stove projects anymore. When I hear Dean say that one millimeter can make a nontrivial difference, it’s inconceivable to me that all these local stovemakers can make all these stoves efficiently. You have to work in a different way.”

GlobalWarmingBox1 To Build a (Better) FireThree years ago, on a taxi ride in southern China, Still had a glimpse of the future. He was working as a consultant for the EPA at the time, passing through the city of Kunming, when he spotted some odd little stoves for sale on a street corner. He shouted for the driver to stop and stepped outside to examine one. “It was like Shangri-La,” he told me. The stove was meant for burning coal, so its design was all wrong for wood, but it was sturdy, compact, and cleanly manufactured. More important, its combustion chamber was made of a hard yet miraculously light and porous clay—a combination that stovemakers had been scouring the earth to find. “There, in this two-dollar coal burner, was everything needed to make the world’s perfect rocket stove,” Still says.

The stove had a telephone number printed on it, so Still called it on his cell phone. Two months later, he was visiting the factory in eastern China where the stove was built. Within two years, the factory was producing a stove to Aprovecho’s specifications. Sold under the name StoveTec, it isn’t much to look at: a hollow clay tube clad in green sheet metal with an opening in front and a pot support on top. But it incorporates all ten rocket-design principles with a consistency that only mass production can offer. The StoveTec uses about half as much wood as an open fire, produces less than half as much smoke, and sells for eight dollars wholesale. In the U.S., where it retails for five times as much, it has been especially popular among Mormons and survivalists.

Still’s stove is a kind of proof of principle. It shows that an efficient, user-friendly stove can be mass-produced at a cost that even the very poor can afford. But it also shows what’s missing. The StoveTec isn’t suited to some dishes—tortillas, chapatis, heavy porridges—and its life expectancy is less than two years. While it’s much less smoky than an open fire, it can’t quite meet Still’s six criteria.

The search for the perfect stove continues, in other words. Not long before Stove Camp, I visited an organization called Envirofit in Fort Collins, Colorado. Envirofit’s laboratories are housed at Colorado State University in a converted power plant from the 1930s. On the morning of my tour, half a dozen experiments were going on simultaneously. One glass case held nine stoves, all furiously burning pellets fed to them by an automatic hopper. Across the room, the smoke was being parsed into its chemical components by a rack of blinking machinery. (Wood smoke may not be cyanide, as Still put it, but hydrogen cyanide turns out to be one of its trace elements.) On a catwalk upstairs, a programmer was modeling green-and-yellow flames on his computer while a biologist down the hall was subjecting live human lung cells to wood smoke. “We grow them in the basement, but they’re fully functional,” I was told. “They even produce phlegm.”

Envirofit’s CEO, Ron Bills, is a former executive at Segway, Yamaha, and Bombardier. His new company is technically a nonprofit, yet Bills believes that, for too long, stovemakers have treated the developing world as a charity ward instead of a business opportunity. “A lot of the poor—call them emerging consumers—get inundated with crummy stuff,” he told me. “So we’re going back to Henry Ford.” Envirofit’s first new product was essentially a rebranded version of Aprovecho’s stove, made by the same Chinese factory with a few improvements in durability and design. In July, however, the company unveiled a new model. It was shaped like an ordinary rocket stove, though much more stylish, and had a major innovation at its core: a durable metal combustion chamber. Made of an alloy developed together with Oak Ridge National Laboratory in Tennessee, it could withstand the caustic fumes of a wood fire for more than five years, yet cost only three dollars a unit to produce. The Envirofit combustion chamber could be shipped for a fraction of the cost of a fully built stove and adapted to local designs and cooking traditions. It was mass production and appropriate technology rolled into one.

“That’s the goose that laid the golden egg right there,” Bills told me. “That’s the Intel inside.” He had nothing against groups like Aprovecho, he said. They could continue to hold their Stove Camps and sell their stoves made out of clay. “But Henry Ford didn’t stop with the Model T. If we are going to make an impact in my lifetime, it has to be done at scale. And when you have a 3 billion–product opportunity, what is enough scale? One million, 2 million, 5 million? I like to dream big.” Thanks, hippies, he seemed to be saying. Now, please step aside.

On the last day of Stove Camp, I stumbled out of bed late, in search of coffee—a timber train having catapulted me awake, as usual, four hours earlier. Aprovecho was as busy as a science fair. The pulmonologist from the National Institutes of Health was putting the finishing touches on a rocket stove made from an oil drum. A Norwegian designer was running emissions tests on a little tin gasifier. And another camper was watching emission measurements unspool across a laptop. “Look at that!” he shouted. “It’s flat-lining! There’s almost no particulate matter!” On the whiteboard next door, the words “Save the World” had long since been erased and replaced with mathematical equations.

Scott and Andreatta were in a far corner of the workshop, probing their injera stove with an infrared thermometer. Their week had seen a succession of setbacks and breakthroughs. When their first prototype, with its steel griddle, had had too many hot spots, Scott suggested that they try aluminum. It conducted heat even better than steel and was considerably cheaper. A few emails to Ethiopia had confirmed that the metal could be locally cast from recycled engine blocks. By the next morning, Andreatta had roughed out a plywood mold for the griddle and they’d taken it to a foundry in Eugene. But the design proved too complicated to cast—it had radiating fins along the bottom to distribute the heat. So they’d settled on something simpler.

The new griddle was one-third of an inch thick and flat on both sides. Andreatta had put a ceramic baffle beneath it to temper and diffuse the flames, but he still had his doubts. The melting point of aluminum is 1220 degrees Fahrenheit—about half as high as the peak temperature inside a rocket stove. If they weren’t careful, the griddle would dissolve before their eyes. Andreatta switched on his LED headlamp and peered at the infrared thermometer. For now, the griddle was holding steady at 433 degrees—just five degrees short of the target temperature. Better yet, the center was less than 25 degrees hotter than the outer edge. “Even Ethiopian women don’t get it in that range,” Scott said.

Still strolled by, wearing a T-shirt with a giant longhorn beetle on it. He had a groggy grin on his face, as if he’d just woken up to a redeemed and revitalized world. Sometimes he saw the stove community more as Ron Bills seemed to see it—as a gathering of undisciplined hobbyists engaged in the equivalent of building iPods out of toothpicks and aluminum foil. But this wasn’t one of those days. Earlier that summer, a research group under Vijay Modi, a professor of mechanical engineering at Columbia University, had surveyed cooks in Uganda and Tanzania who had tested a variety of improved stoves. In both studies, the StoveTec/Envirofit design had won the highest rating, beating out the most recent Envirofit stove in the Tanzanian study. “My people, they aren’t always very smart,” Still had told me. But they were inventive, resourceful, and doggedly resilient. And, after 30 years of trial and error and endless field research, they understood fire very, very well.

The injera stove was the kind of project that might always fall to them. “What is the market for an improved cookstove, really?” Still said. “People hope that it’s big, but we have an eight-dollar stove and it’s not easy to sell. Everyone forgets that poor people are really poor.” In Africa, where less than a quarter of the population has electricity and the most efficient technologies are beyond reach, an open fire can still seem hard to beat, if only because it’s free. “But you know what? We’re going to do it,” Still said. “A lot of people think that if you don’t make a whole lot of money at something, it can’t be good. I think those people are wrong. If you want to do what poor people need, and you really don’t stop, you’re not going to be rich. Not unless you’re a lot smarter than I am.”

Just before we broke camp the next morning, Scott came to find me in the meat locker: the prototype was ready for its first pancake. He and Andreatta had hoped to cook true injera bread for the occasion, but they couldn’t find the time—or the teff—to make a proper sourdough. So they’d settled for Aunt Jemima. “This is our first test,” Scott said, holding up a pitcher of pancake batter. “People of the world, cut us some slack.” Then he poured it onto the hot griddle.

Over the next three months, the stove would go through more rounds of fiddling and redesign. The aluminum would prove too conductive for real injera and get swapped out for a traditional mitad. To get the ceramic to heat evenly, the baffles beneath it would have to be removed. At one point, in Addis Ababa, Scott would nearly abandon the project, only to have an Ethiopian cook make some key suggestions. Yet the result would be even better than it seemed on this sunny August morning: the world’s first successful rocket injera stove—twice as efficient and many times more durable than those it was meant to replace.

As the batter hit the griddle, it spread into a circle that nearly reached the edge. Within a minute, it was bubbling up evenly across its surface. “Yeah, baby!” Scott said. “If we’d tried that last Friday, it would be blackened char in the middle.” He slid a spatula under the batter and tried to flip it, leaving half on the griddle but the rest well browned. He stared at the pancake. “We can’t really fucking believe it,” he said. “I mean, these designs usually take months and you’re still scratching your head.” The stove was almost ready, he thought. Now they just had to convince a few million Ethiopians. ♣

Burkhard Bilger’s work has appeared in The Atlantic Monthly, Harper’s, The New York Times, and numerous other publications. Bilger is now a staff writer at The New Yorker. He served as senior editor at Discover from 1999 to 2005, and before that he worked as a writer and deputy editor for The Sciences. His book, Noodling for Flatheads: Moonshine, Monster Catfish, and Other Southern Comforts was a finalist for the PEN/Martha Albrand Award for First Nonfiction.

Adapted from: “Hearth Surgery” By Burkhard Bilger.
The New Yorker
, Dec. 21, 2009

Illustration by Dan Page



  • Kip Hansen June 24, 2011 at 5:22 pm

    If you want to give your body back to nature, it should be buried without embalming, not enclosed in a plastic bag, and in a bio-degradable box. Alternately, burial at sea, again without poisonous embalming fluids,would do the trick. It takes a lot of ingenuity to truly give that body back. Good luck.


  • Astrid Caldas: Arctic Ice, Polar Bears and... Soot? August 2, 2011 at 10:43 am

    […] inefficient cook stoves with cleaner burning ones. Recently, Conservation Magazine published an article on the efforts to create a cleaner-burning stove that would be affordable to the communities in […]


  • Christopher Yang February 27, 2013 at 10:40 am

    What was the materials and design of the stove that Still found in China?


  • Elly Grogan March 6, 2013 at 11:56 am

    Those of us at StoveTeam International are encouraged that Conservation Magazine is covering the work us ‘stovers’ are doing to address the incredible need for improved cookstoves throughout the world. StoveTeam International is also located in the Eugene area, and our model is a bit different. We help entrepreneurs in developing countries to create their own stove manufacturing facilities that are independently owned and operated. The Ecocina stove we support for Latin America is made of all locally available materials – pumice, cement and field tile, and stoves that have been in the field for four years are still viable. Some of this is due to the local manufacture which can easily adapt to changes recommended by local users.


Leave a Comment