Editor's Note: Students from Dartmouth's Thayer School of Engineering are working in Tanzania to help improve sanitation and energy technologies in local villages. This series chronicles work being done by the student-led group, known as Humanitarian Engineering Leadership Projects (HELP), to design "rocket stoves" in the village of Mwamgongo and top-light updraft design (TLUD) gasification stoves in the village of Kalinzi. The goal is to create a healthier, more energy-efficient cooking apparatus that these villagers will accept and use. HELP students are filing these dispatches from the field during their trip. This is their eighth blog post for Scientific American.

Time crawls from morning to night, but the familiarity of each day blurs one week into the next.  This is unfortunate because I have a stove to build. "Build" is actually the wrong word. We built stoves last year. This year we're going to design a stove.

There is no set design process, no formula to follow, but what is important is understanding what the user wants. This means it's time to talk to the experts: Women who spend up to seven hours a day cooking for their families in smoke filled kitchens. Last year we built stoves for a pilot group of 25 households, which would be the basis for design feedback this year. I grabbed my pen and notebook and left the field station eager to see some familiar faces and hear their impressions of the stove.

Running to greet us as we left the field station was a young boy named Kilai. He ran over with a huge smile on his face spewing greetings in Swahili and a few broken phrases we had taught him in English last year. “Good Evening,” he beamed. I checked my watch, yup, it was 8:30 a.m. Kilai is probably about 10 years old (he doesn't actually know) and is my favorite person in Mwamgongo. His life story reflects some of the shortcomings of the educational system in Tanzania. Kilai caught polio as a child, and it left him with serious learning disabilities. Because the school in Mwamgongo has no resources for special needs students, Kilai dropped out of school at around nine years old unable to read, write or perform simple arithmetic. Kilai spends his days hanging around our house, waiting until he is old enough to become a fisherman. What he lacks in formal education, he makes up for in personality. Kilai is loved by everyone in the village and has appointed himself our "side-kick."

With Kilai in the lead, we set off to survey the village. What I saw in the field was disheartening. People were not using the rocket stoves, but, thankfully, they had many complaints.  The primary problem was that every stove we observed was damaged in some way. The main difficulty was that the "top knobs"—balled up pieces of clay that elevate the pot off of the stove permitting airflow into the combustion chamber—had broken off. This limited the flow of oxygen, which inhibited combustion. The result was burning stoves that smoked a lot. People also complained that the firewood intake was too small to accommodate their firewood and that the sticks needed to be pushed in every few minutes, which made it difficult to walk away from the stove. We also observed that the "superstructure" (surrounding mud cement bricks) of many stoves had broken, severely weakening the stoves. It was depressing to see such poor results after the first year, but I understand that failure is part of any design process and the feedback we got will certainly inspire a new design.

I took what we learned from the test group and put it into a specifications table. A specification is a characteristic of the stove that can be evaluated. By standardizing the criteria you use to evaluate the stove, you can accurately compare the performance of different designs. For example, the time the stove takes to boil one liter of water is a specification of the stove that can be measured (in minutes). Thus, we can use "boil time" as a point of comparison between different designs. Moreover, from user group surveying, minimum values for the specifications can be determined. One example is that users may demand that the stove boil one liter of water in less than 10 minutes. The design of the stove would then need to meet that quantitative value in order to appeal to the users. The specification method allows for formalized testing and reduces the ambiguity when comparing designs.



Justification for Specification

Metric of Evaluation (based on user group surveying)

Cooking Performance

Stove must cook traditional meals

User group testing. Binary yes/no

Boil Time

Stove must cook food at a pace that is acceptable to the user

Boil time for 1 liter of water must be < 10 mins, 3-stone stove boils in 7-10 minutes


Stove must be as efficient as possible to reduce fuel wood consumption

At least 20 percent (twice as efficient as 3-stone stove)


Must be affordable by all people

Must cost less than 3000Tsh per stove (< 2 US dollars)


Stove must last for a long time with limited maintenance

Stove should last 2 years without maintenance

Ease of Construction

Stove must be easy to construct and easy to repair

Must use locally available materials and building techniques. Building time including materials gathered < 10 hours. Cost of maintenance after two years < $1 time of maintenance should be < 5 hours


Stove should be visually appealing

Does user group approve? Binary Y/N

Burn Protection

Stove should prevent against burns

After cooking cycle (2 hours) outside superstructure of stove should not be able to burn. Binary Y/N

Indoor air pollution

Particulates should be reduced, CO emissions should be reduced

Testing with pariculate and CO monitors. More research must be done to find acceptable levels of emissions


Stove should be sized to fit in existing kitchens and should not be too tall (women are used to working on short stoves)

Maximum height 40 cm

Maximum width 60 cm

Maximum length 60 cm

Firewood size

Must be able to accept firewood that people normally use

Must accept firewood sticks that are 12 cm wide

Burn time

People like to put wood in their stove and walk away. Stove must be able to burn for an acceptable period of time between wood additions

Must burn for longer than 5 maintenance free minutes


The values for this chart were derived from extensive discussion with the 25-sample user group. Hopefully, this chart gives some insight into the criteria we're using to evaluate the stoves. At this point, we need to start designing. If these specifications are unattainable then my time will be limited in Mwamgongo but I'm confident we can meet most of these specs and will have a successful stove, it's just going to require some serious work.

Image courtesy of Wendy Hado