The Tega: An Inexpensive and Small Solar-Thermoelectric Generator

By George Geng and Preston Zhou, THINK 2013 Winners


Our project idea was to make a solar-thermoelectric generator, called the Tega. This uses a temperature gradient between two sides: one heated by the sun and another immersed in water to keep it cool. On the top side, we had a solar cooker oven that concentrated sunlight onto a small area. On the bottom side, we had a metal plate that conducted heat and dissipated it in water. During the day, the temperature difference between the top and bottom plates can generate power. During the night, the water can cool the Tega down. Overall, we wanted to create an inexpensive but viable alternative to conventional solar panels.


We started by constructing a basic solar reflector, except it didn't produce enough heat. We used a laser pointer to see where the light reflected to when the sunlight hit, and most of it missed the hot plate on the top side.

So, we made modifications to the original design. We put a box covered with Plexiglas below the solar reflector, to keep the hot plate from cooling down by air convection. We improved the solar reflector as well, by calculating the optimal setup, and reached our temperature goal with 350 °F (177 °C).

However, once we tested them with the Peltier plates, we found that they generated almost no power. The temperature gradient gave us only 0.04 V and 45 mA. It turned out that the Peltier plates should have been used to create the gradient, instead of the other way around. After some research, we discovered that a 150 °C difference should give 8 V, so we bought a new set of thermoelectric plates.

We tested them with the existing design, but the newer plates only generated 0.3 V. This made us wonder if the problem was with the design, not the plates. So we put a stove on the top side and an ice pack on the bottom side, which generated a maximum of 0.5 V. We believe that there is a small design flaw that prevented the plates from working completely. For example, some Peltier plates require pressure to work. We are continuing to troubleshoot the plates, in hopes of getting more voltage and power, and experimenting with a large Fresnel lens for our solar cooker.


Our project goals were to (1) get a temperature gradient of 100 °C and (2) generate 30 W. At this point, we achieved and surpassed the first part. A large temperature gradient is crucial for powering the Tega, and we established it using relatively simple, everyday materials like cardboard and duct tape. In our testing in May, the hot plate in the Plexiglas-covered box quickly reached scorching temperatures of 149 – 177 °C (300 – 350 °F). Accounting for the cooling water, which was only 28 °C (82 °F), the temperature difference was up to 150 °C.

According to the Peltier plate specification, we should have gotten 8.7 W per plate. Since the Tega had 6, it should have produced 52 W. Compared to the theoretical output, our Tega prototype only produced 0.0255 W. Since it didn't achieve the intended power output, we are troubleshooting this issue over the summer (trying to apply more pressure and set it up anew in order to make it work). Our hypothesis is that the issue is the intermittent contact between the two sheets of metal and the opposing plates of the Peltier plate. As a result, it is not transferring enough heat, and the two sides of the plate are not reaching the expected temperatures. We plan to amend this problem, finish troubleshooting the entire design, and reach our intended goals in the next year.

We found that an idea is not easy to bring to life. For the first prototype, we found that our simple design was very tricky to implement and get right. Some of the mechanical issues we ran into were how to clamp the thermoelectric plates to each side and how to account for tracking. This project taught us the difficulties of thermoelectrics; we had a hint from reading related journal articles, but we fully realized how challenging it is for real-world power generation during the implementation process. Although we did not meet all the goals we set, we plan on building on our progress and have the Tega running soon.


Our Tega design achieved our first goal: the temperature gradient, which can be difficult to maintain and is essential for thermoelectric power generation, worked better than we expected. We should have gotten enough power to be competitive with traditional solar, according to the data sheet. But the thermoelectric plates did not work with the stove either, so we need to focus on them. Luckily, this gives us a clear direction for the next year of this project. We will continuously improve upon our design and eventually, make the Tega reach our intended objectives.


We thank the THINK team, especially Somak for supporting us and organizing the event and Youyang for hosting us at MIT in February. The team made the THINK program a wonderful experience and opportunity for helping us develop an idea by ourselves. We specially thank Thomson Reuters and Chevron for their generosity, which allowed us to embark on this endeavor.