ASK THE EXPERTS: Thermoelectric Generators

A Tecteg thermoelectric module—at 2.2 square inches, it can make between 5.5 W and 15.6 W at temperature differentials between 100°F and 210°F.

My off-grid house uses a gas generator for primary power. I charge a large pack of Iron Edison batteries and have a 4 kW inverter. I also have six PV modules that charge the battery pack, but they produce negligible energy in the cloudy, short daylight hours of winter.

I am interested in designing a thermoelectric generator (TEG) system to reduce my generator run-time in winter. My idea is to use a combination of solar water collectors and a wood heater for the heat source, and a deep well as the cooling source—and maybe also heat my greenhouse with waste heat. Can you point me to a book or a website where I can get help designing my TEG-based system?

Ric Barline • via email

TEGs are fascinating cousins of photovoltaic (PV) cells, and preceded the discovery of the PV effect by almost 20 years. TEGs must have a temperature differential (∆T) to function; one side must be heated and the other side cooled.

The most glamorous use of TEGs today is in spacecraft traveling so far from our sun that PV arrays won’t work. The heat is provided by the decay of radioisotopes, such as plutonium, strontium, and others. TEGs have also been used to power remote military installations, and today are used for recovering waste heat from industrial processes. Automotive TEGs are being developed to recover part of the large amount of engine power (~30%) that is wasted as tailpipe heat.

The biggest drawbacks of TEGs—especially for an off-grid home application using wood heat—are efficiency, cost, and reliability. Commercial wood heater TEGs run $10 to $20 per watt, and are only 5% efficient. Compare this to PV modules that cost less than $1 per watt and are 20% efficient. Manufacturers of wood heater TEGs have come and gone. The biggest problem is TEG burnout—if your heater is running hot and the cooling on the other side of the TEG can’t keep up, the unit can be ruined.

There are a few manufacturers of wood heater TEGs, but the limitations are many. The information for DIY units focuses on small applications, like propane camp stoves that can charge a smartphone or run a small LED light. I did the math for my wood heater. Covering all the usable surfaces with TEGs would cost about $3,000, and produce less than 300 watts. And cooling the other side of the TEG would be a challenge. I’m going to stick with PV and a small, reliable backup generator. If your project is a success, though, please let us know. The TEG concept does have potential in certain applications, and many people will be interested in your findings.

Dan Fink • Buckville Energy Consulting

Comments (2)

Ed Hutchinson's picture

I too am fascinated by TEG devices. It seems to me that a device that is primarily designed for TEG, rather than TEGs attached to a "heater" may help. I think a (tiny) wood pellet (or gas) burner with lots of sensors and temperature controls might work. If the 94% of heat that has to pass through a TEG (to make 6% of the heat into electricity) is used to heat space or hot water then the efficiency doesn't matter. So a device producing 500W running 24 -7 will produce 12kWh per day, and dissipate over 8kW of heat (>190kWh) per day. That is a lot of heat. Combined with enough PV, and a day or two of battery storage, one would think the TEG would get an off-grid system through the winter without running a generator at night. With retail TEG modules at $2.37/watt (I bought some today), I hope to build a small "boiler" for about $4000, possibly less. That would cost less per watt than my first kW of PV (2008) and that PV system didn't heat the house. The PV system also only generates about 1200 hours per year (equivalent at full power), and a boiler running 24-7 for 90 days would generate 2160 hours per year. So for the cost per watt one gets almost double the output per watt of capacity.

gerard's picture

Hi Ed,
Not easy to build what you are building wish you luck. We have been in the thermoelectric industry for the past 25 years. Over the last 3 years we have been designing a Fully integrated system with in put sensors and output sensors to control and number of variables important in correct operation of a thermoelectric system. No easy task I might say. The result has been the soon to be launched Rabbit Ear thermoelectric generator for wood stoves. The unit will attach to the flue just above the stove and penetrate the flue to absorb direct heat from the gases passing bye. This will provide ample heat flux absolutely required to feed the TEG modules to operate at close to full capacity. The system has a custom brain Charge/ Load Controller with PLC functionality. You can read up on it if you search rabbit ears thermoelectric Generator.

Show or Hide All Comments