Thermoelectric Generator Shines Where The Sun Doesn’t

For off-grid renewable electricity, solar seems to make sense. Just throw some PV panels on the roof and you’re all set to stick it to the man, right? But the dirty little secret of the king of clean energy is that very few places on the planet get the sort of sunshine needed to make residential PV panels worth their installation cost in the short term, and the long-term value proposition isn’t very good either.

The drearier places on the planet might benefit from this high-power thermoelectric generator (TEG) developed and tested by [TegwynTwmffat] for use on a wood burning stove. The TEG modules [Tegwyn] used are commercially available and rated at 14.4 volts and 20 watts each. He wisely started his experiments with a single module; the video below shows the development of that prototype. The bulk of the work with TEGs is keeping the cold side of the module at a low enough temperature for decent performance, since the modules work better the higher the difference in temperature is across the module. A finned heatsink and a fan wouldn’t cut it for this application, so a water-cooled block was built to pump away the heat. A successful test led to scaling the generator up to 10 modules with a very impressive heatsink, which produced about 120 watts. Pretty good, but we wonder if some easy gains in performance would have come from using heat sink compound on the module surfaces.

Using thermal differences to generate electricity is nothing new, but a twist on the technique is getting attention lately as a potential clean energy source. And who knows? Maybe [TegwynTwmffat]’s or one of the other Hackaday Prize 2018 entries will break new ground and change the world. What’s your big idea?

81 thoughts on “Thermoelectric Generator Shines Where The Sun Doesn’t

  1. I find it extremely hard to believe that a peltier generator on a wood-burning stove is going to be either economical or renewably sustainable in any kind of large-scale application. Especially compared to solar with its rapidly-dropping prices.

    I love playing with peltiers, though. The heat sink work is nifty.

        1. Yes, why not? Burning wood releases carbon which was captured in recent years as it grew and which would be released soon anyway if allowed to decay. CO2 from very old ‘fossil fuels’ is releasing carbon trapped millenia ago adding it our modern atmosphere. We should avoid that.

      1. Commercially viable for what? We use TEGs to power small wireless sensors at work where it becomes expensive to access the device every year to replace a battery. Works really well and paid for itself in one year.

    1. Solar is a borderline boondoggle. Max production is at solar noon and max usage is in the morning and evening. Batteries still suck and just do the math on the cost of a battery that can store 12 hours of production of a moderately sized power plant.
      Wind is a much better solution for alternative energy and the costs are much lower. Plus while not demand following they at least sometimes produce power at a peak demand and at night.

      1. Statistically, winds follow the day/night transition, so the strongest winds happen some hours after sunset when the ground begins to cool down and the air density starts to increase while the previously heated air is still rising up; end result is a strong sideways flows of air to fill the resulting low pressure zone.

        In Texas for example, the wind tubines produce most often during the late evening or early morning when the power demand is the lowest. The power utilities are even offering free electricity in the night to get the people to use it, and the main reason for doing so is because federal subsidies are tied to the amount of energy produced.

        Renewable energy has very little to do with producing clean energy, and a lot to do with getting tax breaks and free money.

        1. Luke: “Renewable energy has very little to do with producing clean energy, and a lot to do with getting tax breaks and free money.”

          Not true. The *purpose* is to get energy without adding CO2 to the atmosphere (and ideally without nuclear, but that’s a lot to ask!) while reducing dependence on the Middle East for Oil.

          If subsidies are offered to help this happen, then good! (Some will cheat and manipulate the system and this needs to be minimised and loopholes closed.)

  2. Awesome. I know these devices are not all that efficient as I have several over here that I play with from time to time but, I do believe they will have their place and, as time goes on, folks come up with new and creative ways to deploy them,

  3. I have thought about the same thing a year ago but I don’t have a stove for that. The idea is sound. A bad efficiency is not a problem. When using a stove, you want to generate heat (huge amounts of it in terms of energy in Watt) anyway. A small fraction of that in electricity instead of heat should be sufficient to satisfy the demand for electricity. Also, the stove is typically used when the sun is not shining or where the sun is not shining often. In that aspect, this is a good counterpart for photo-electric/solar

  4. Your claim that there “few places” where there is enough sun to make PV worthwhile is just plain wrong and burning wood to run thermal electric is very likely not a clean or efficient solution.

          1. Don’t put the heat exchangers in the chimney. You can put the heat exchangers on the stove. That is the right way to do it. As I wrote above, in comparison to the heat that you WANT to generate, the required electricity is only a fraction of that. That is why those inefficient modules could be enough. That is why you really can effort it to generate electricity instead of heat out of the wood. Then, what do these modules do, being so inefficient? They “produce” heat (in terms of release heat). But that’s what you wanted to do anyway when using the stove.
            Then, what do you do with the electricity? Running the computer, switch on the light? This is all converted to heat that you wanted to generate anyway when using the stove. So, when someone says, this approach is not clean or efficient, what is he/she talking about? ;)

          2. and to add to that you can distribute the heat better by locating the radiators in another part of the house. part of the downside of using a wood stove is that it usually only keeps the room that its in hot.

    1. Burning wood can be clean and efficient, if a rocket stove design is used. (i.e., in less than ten minutes, they stop smoking as the chimney -which can top 1000C- passes the auto-ignition temperature of everything in it.) The temperature difference is so extreme, I’m wondering if a Stirling engine wouldn’t be more efficient than a Peltier solution.

      1. >The temperature difference is so extreme, I’m wondering if a Stirling engine wouldn’t be more efficient than a Peltier solution.
        Now THAT’s interesting. I hope someone has numbers.

        A friend’s well hole also bubbles up water year round at 4C. There’s his cold source. Air in winter is even colder.

      2. Stirling has moving parts, is not maintenance-free, requires more space, makes noise, is harder to build/acquire and is harder to mount on a stove. Also, it would be appreciated, if the system could work with lower temperatures. The fire can ans should be hot. But you don’t want the outside walls of your stove at 1000C,.

        1. Have to agree.
          The seebeck (article wrongly refers to peltier) generator is good, but I would say the efficency is much lower than either sterling or turbine. I think even the theoretical maximum efficency of seebeck is lower than what sterling/turbine engines *currenty* get.

          Complaints about moving parts? True, but we’re on earth, not in space, where things with moving parts are a nightmare. Machines with moving parts have been used for centuries.

      1. It is economical for the national economy as soon as you tell me where the electricity comes from in winter time. Solar is economical for the owner in many cases due to massive direct and indirect subsidies. As a consequence among other things, Germany has a expensive distribution network with massively increasing costs and decreasing quality in terms of availability. For a private person, one kilowatt hour costs almost 40 US cents including all kind of taxes! In those 40 US cents per kwh, the amount of fossil fuel (coal, oil) is basically non-existent nowadays. Still, the part of the TOTAL amount of Germanys energy that is covered by solar and wind is approx. five percent. Let me make this clear:
        Approx. five percent of Germanys TOTAL energy requirements are covered by solar and wind. This is the annual average. The number is much smaller in winter time. And Germany already has a lot of solar and wind.

        1. Well this is clearly wrong…

          In march 2018 the wind share of total electricity consumption in Germany was 21.6%, solar accounted for 5.9%. Total renewable share was 40%.

          Comparable numbers for all of 2017 was wind 18.8%, solar 7%, total renewables 38.3%.

          1. You forget direct consumption vs. nominal. Germany exports much of the wind/solar power to other countries because it cannot be used locally. The difference between peak and average output is just too great for any system operator to integrate the power on the scale of a single European country.

            If you take the whole EU synchronized grid that countries like Germany and Denmark use as their “virtual battery”, the true amount of solar+wind in the system drops to about 6-8%

          2. I wrote “total energy”, not “total electricity”. Don’t twist my words and get your facts right. There is a lot of other demand that is (currently) not electric. All the effort that Germany does and all this costs and the expensive energy are really only about five percent. The other 95% percent still have to do with burning stuff, for instance by clearing forests in Eastern Europe. Thanks to the EU, this does not count as wood imports.

            If you understand German, here is the Wikipedia:
            PV, wind and water was 3.9% in 2016. So subtract water from that, which makes this number smaller. Consider, some other factors and you end up at approx. five percent (rounded up). This is annual average, not winter time.

            I wrote about PV and wind. Germany always had always had water. Water does not increase the cost of energy like PV and wind does. Also, bio diesel isn’t the cost-pusher.

            Germany spent and spends dozens of billions of Euros (US billion) on solar and wind which really only make for less than five percent of the total energy!

        2. There are times when Germany produces so much electricity from Wind and Solar and the demand is lower than the supply that their electricity prices invert into the negative range.
          During those times, significantly high percentages of their electricity mix comes from renewables.

          And according to Fortune, enough renewable energy was produced to cover 32% of their energy use in 2016. Not 5%.

          Considering they are still investing in the technology, that was probably a higher percentage in 2017 and will be even higher this year. Actually, seeing Preben Dyrholm’s message, I see that was right.
          We have much better options available to us than consumable sources like coal, oil, natural gas, or nuclear. It’s not as expensive as the naysayers claim either. And it benefits the country in the long run, even the naysayers.

          1. > And according to Fortune, enough renewable energy was produced to cover 32% of their energy use in 2016. Not 5%.

            This is obviously wrong. It can’t be both, approx 30% of electrical as well as approx 30% total. Let’s stay with the official numbers and lets not use numbers from idealistic journalists that write like they would like the world to be. That is different from reality.

            And again, I was talking about PV and wind. PV and wind are less than 5 percent of the total demand. Look it up!
            Google Translator:
            -> 4.1 Primary and final energy consumption
            -> First table in subsection 4.1

            renewables in 2016:
            Fuels from renewable energy sources: 8.7%
            Photovoltaic, hydro and wind power: 3.9%

            Fuels from renewable energy sources will include wood from Eastern Europe and oil from areas previously covered with rain forest.

            Also have look at the other numbers in this table.

            Yes, Germany has some hydro. Yes, Germany produces some bio diesel/gas. Yes, Germany burns wood (nowadays a lot of it). This has nothing to do with the costs of the German “Energiewende” (Energy transition).

            By the way, it is so great to have a couple of a additional electric kwh when you DONT NEED them. You know that you can’t have those for a reduced price, if you are a German? Those are sold to other countries, for a negative price, if required. And guess, who pays selling those kwh with a negative price.

      2. Only economical once the price is raised to subsidize the cost, Germany’s average KWH price is almost 5 times higher then what I’m paying in the Midwest US even with clean energy surcharges.

        1. That’s not really true. Panel prices are $1-2 per watt these days, so a 100w panel will run you between $100-$200, retail. Even at $0.11/kwh it won’t take an extremely long time for those panels to pay you back. You reap additional benefits as well, such as having a local backup in case the grid is down. Not to mention it doesn’t require a consumable to generate electricity and it doesn’t pollute while doing so either.
          Solar panels are much cheaper and more efficient today than they were even a few short years ago.

  5. Don’t get the peltier too hot, or it will de-solder itself internally. There’s a maximum operating temperature given in the data sheet, if available. Common peltier elements are only rated up to 200 C because they’re properly meant for cooling things instead of generating power, and you need a specifically made TEG element if you want higher temperatures. They come in different grades up to 600 C max operating temperature – around the melting point of aluminium.

    The irony is that semiconductors lose properties in the hot, so the peltier element works the worse the hotter you get it.

    1. Youre wrong.
      They work but the temperatur differece needs to be large.

      And you are right about the desolder temp.
      Going above 180C and things start to get apart.

      But keeping the temps bellow and it will work great.

      1. >”Youre wrong.”

        About what? The whole point of a TEG is that you got two opposing semiconducting materials side by side. One where the electrons are the main charge carriers, and another where the electron vacancies of holes shif the charge. Heat makes the material vibrate, which causes the charge carriers to act a bit like gas: heated up they expand and cooled down they condense.

        So, with one end of the material heated up and the other cooled down, you get a density difference of charge carriers, and therefore a difference in voltage. Two different types of semiconductors create opposite polarities for the + and – charge carriers, and when you have these two side-by-side, it creates an infinite loop that carries electrons around for as long as the heat is flowing through both materials.

        However, semiconducting materials become more like regular conducting materials when the temperature goes up, so getting the TEG hotter makes its efficiency drop because the main effect responsible for the current gets weaker and weaker. For silicon based materials the limit is around 125 C. That’s why your CPU starts crashing when you overheat it – the transistors just become resistors instead of switching anything.

        1. At higher temperatures a Magnetohydrodynamics generator might be a better choice. Actually, I would think that if you’re going to use combustion to generate electricity and don’t want to use a steam driven turbine or sterling engine, an MHD would be a better choice. And the exhaust gases from an MHD are hot enough that you could still heat a boiler to drive a turbine and extract further energy.

          I think another problem with the peltier is you need not only a hot side but a cold, or cooler side and it’s this temperature delta that is important. The problem is the device itself isn’t the greatest insulator and you probably lose a lot or energy through leakage.

  6. Weather stations near the poles use thermo-electric power from propane flame to power its electronics. The “waste heat” keeps the station from becoming a block of ice. And a tank of gas lasts a year. Of course, the big Mars rover and both Voyager spacecraft are TC powered from Thorium heat ‘Seems that TG has its uses here and there.

    1. Space probes use Plutonium 238 exclusively. Thorium would not be suitable for this application…

      You need something that undergoes alpha decay and has a half-life somewhere between 50 to 100 years. (shorter would mean the generator loosing power too fast and longer would mean a lot more material needed for desired output power)
      Also daughter products should not decay with high-energy gamma emissions, as those are considerably more difficult to shield.

    1. Yeah – that first paragraph is very down on solar. Made me wonder if the author was a climate change denier!

      I watched a documentary in 2016 that claimed that it had become cheaper in the Arab Emirates to use solar power to pump oil/gas out of the ground. It doesn’t get much closer to the source than that, so if it is more expensive to use oil/gas at the point of extraction…
      For the interested, the documentary is on youtube at about 50 minutes.
      Then again, you can’t believe everything you see or hear. And that includes stuff on hackaday.

  7. Anything NASA sends into deep space uses the hot side cold side to generate current. The hot side is a ISOTOPE, keeps the gears and electronics warm. The current is a supplement to the solar panels. Which far away from the sun don’t generate much. It’s a win-win.

    1. RTG’s have been around for a long time. They used them in the USSR to power remote lighthouses and after the country fell apart, people saw them as great sources of aluminum scrap. I guess they did not believe that they were actually full of radioactive stuff. They used to use small versions in pacemakers too.

      1. There’s a site with all the reports of radiation caused injuries and deaths, from the SL1 to Chernobyl to somewhere in South America where a whole village was contaminated by people spreading around the pretty blue glowing Cesium powder from an improperly scrapped medical imaging machine.

        It also lists all the (known) incidents of people stealing RTGs from Soviet lighthouses and peeling off the shielding to use them as heaters for their hunting and fishing camps.

        Build a man a fire and he’s warm for a day. Warm him up with an unshielded RTG and he’s warm for life.

  8. Wow. Dan, you’re my hero today.

    I’ve had comments deleted just for pointing out that solar is a worse financial alternative than investing pre-tax money into a conservative 401k… even through 2008..

    So I’m absolutely amazed the HaD editors green-lit such a ‘blasphemous’ article. Next you’ll be saying a $100k Tesla doesn’t make sense for most people or the “$30k” Tesla isn’t a great value. …or that open source hardware doesn’t make sense for most hardware companies.

  9. The original purpose of a wood-burning stove is to generate heat by burning otherwise wasted biomass so in that regard it is more “efficient” in converting waste into a usable product than simply burning the waste to get rid of it. We’ll ignore for the moment the pollution aspect and whether allowing it to decompose to make more biomass is a viable option. Adding a Peltier device to a wood burning stove now takes some of that wasted heat and converts it into a very useful product, electricity, without increasing the pollution generated or creating new biomass. This is now taking something from what was wasted so it is a net increase in efficiency in terms of the fuel used, but the cost of production of the parts, maintenance, time, etc., can exceed the value of any electricity generated. Efficiency then rather depends on what you consider a workable set up. Comparing Peltier panels on wood stoves to solar or hydro is like comparing apples to oranges. They both generate electricity but do it in different ways with different efficiencies of production. I think the Peltier idea is a good one as long as the other “hidden” costs don’t overshadow the benefits obtained down the road.

    1. The thought of an active water cooled system sounds to me like you are taking something that is already very inefficient and making it even more so. On the flip side, you are probably not burning wood unless it is cold out so why not put the cold side outside? Or use heat pipe which requires no pumps and is very efficient to bring the heat to the devices and keep the devices outside.

  10. Hi-Z Technology has high temperature modules designed for power production. No solder is used in the manufacture. Max temp is rated at 250 deg C. It can also be flipped and used as high temperature Peltier. Probably it’s the highest temp Peltier on the market, though they don’t market it as such.

    Definitely use thermal grease on both hot and cold sides. Peak performance, for a given Delta T, is achieved with even temperature across the two surfaces of the module. Higher Delta T improves efficiency. Over temp on the hot side will produce more power, but will begin to oxidize the thermoelectric material, shortening the module life. Colder is better on the cold side until you get to absolute zero, when you’re in uncharted territory.

    They’ve got a white paper with design suggestions.

  11. I have a 1982 1 ton flatbed truck with plenty of room under the bed. I’d like to build a TEG to run the engine exhaust through, with the Peltiers arranged to output ~14 volts DC. If it could output enough, it’d keep the battery charged and drastically reduce the load on the alternator, perhaps even eliminate the need for it.

    The truck doesn’t have air conditioning. I’ve been thinking of using Peltiers sandwiched between two water blocks. Cold side piped to the heater core and hot side to an AC condenser. IIRC, cooling with Peltiers is most efficient when they’re run in parallel at around half their maximum capacity.

    Even if using TEGs to convert waste heat to electricity then using that to power Peltiers plus a small circulating pump can only knock 10F off the temp inside the truck cab, that’d be ‘free’ cooling that wouldn’t put any extra load on the engine.

    Since it’s a 1982 1 ton, none of this would fall under any laws or regulations for emissions. Until sometime in the 1990’s, one ton trucks were totally exempt from emissions regulations, even in California. The reason being that many were sold with a cab and no bed. With the huge variety of possible 3rd party configurations it was impossible to test them all, still is. That’s why Dodge was able to make the Lil’ Red Express and Warlock pickups in the 70’s with engines built like it was the 1960’s.

    1. I’ve also heard of exhaust heat driven absorption refrigerator based air conditioners – or even air conditioners driven off what’s effectively a steam engine that uses the same refrigerant and condenser. Search for “Einstein Refrigerator” for an interesting example. Both this air conditioning application and the generator seem like they’re better served by a mechanical heat engine, whether it’s a Stirling engine, steam engine, or even something more obscure like an Ericsson cycle engine.

    2. Once you consider the cost of all the Peltier and Seebeck (the same thing in reverse) effect devices, you’ll find that butchering an AC system from a wrecked car (or several…probably won’t find a completely working one in just one) and having it filled by a “pro” will be far cheaper even if you account for all the extra fuel the engine will consume because of the AC running.
      Peltiers are terrible at any sort of larger scale cooling, mechanically driven phase-change systems simply win on both efficiency and cost. The only time you use a peltier for cooling is when a phase-change system would be too big for the application.

      And if you thing that lack of moving parts means no maintenance, you’d be sorely mistaken – there’s all kinds of vibration in a car that tends to slowly break things, even when they’re not supposed to move by design ;-)

    3. That’s a novel, to me, and practical idea. I’m working on the 2001 Dodge Ram with a 360 V8.

      I was thinking about and have another A/C electric clutch to use to potentially have the alternator switchable to reduce drag once I get the additional batteries installed if for some reason I needed more power from the alternator, there was an issue with the solar or there was an issue with charging the batteries. I also thought less use of the alternator would be quieter in regards to RFI/EMI though I really need a diesel to have the rig running and be quieter in that respect.

      I’ve also looked into the Dodge Viper or Lincoln radiator fan to get rid of the fan clutch.

      I haven’t looked into removing the A/C, power steering, oil and water pump yet and replacing with electric versions… though the later two would probably be way more a pain in the modification to perform and then again… all this isn’t the best for the vehicle operations to do radio directional finding… though would definitely save on gas money when traveling.

      The electric fan and alternator mods seems relatively easy enough along with the idea of the use of waste heat for energy and am going to look into this idea more since even if I bash guard lower drag skirt the cab on back more… the engine compartment is going to have air flowing through more in hotter environments and cooling fins should be suitable at the peltiers sweet spot for optimal range of operation in a worst case outdoor temperature environment.

  12. I had thought about using a thermoelectric grid under pavement. The grid points would be the union of the two metals. The hot side would be in contact with the pavement while the cold side would be a spike in to the Earth. There is a huge differential temperature there and plenty of pavement.

  13. “For off-grid renewable electricity, solar seems to make sense. Just throw some PV panels on the roof and you’re all set to stick it to the man, right?”

    1. What is letting solar down at the moment is not the efficency of the panels or their cost, it is the storage (battery) they require to provide the output when it is needed.
      Storage does not have to be chemical batteries, it can be pumped hydro. It’s done in Australia, although it is usually coal derived power that ‘replenishes’ or ‘recharges’ the hydro system. But that can change, and it is. Just slowly.

  14. This seems like it could be useful in the tropics/subtropics. Just bury the cold part of it underground, stick the hot part in direct sun, and then connect the two with a thermally insulated medium. Especially if you just need some electricity or to charge a battery to get you through the night.

  15. Time to beat a dead horse, but man I love beating horses.. I wonder if you could pack these modules around the catalytic converter on your car and have an air scoop on the front to funnel large amounts of cool air through the heat sink. I wonder if you would have a net profit after the drag from the wind scoop and the weight of the modules.

  16. Thanks for the video. I’m thinking of building something like this myself. I already have a woodburner to keep me warm – occasionally it keeps me too warm – and solar panels which don’t really do much for me this time of year and 2 kWh of battery storage which is ample for the amount of electricity I use in a day. I keep a kettle on top of the woodburner, from which I transfer a bit of water occasionally to make tea, fill thermos flasks, etc. It should not be too much trouble to put a Seebeck between the stove and the kettle to power my radio and some lights and possibly even my computer. There seem to be loads of different Seebeck/Peltier modules out there. Can you recommend a particular part number?

    I am doing this because I find the UK grid suppliers disappointing in various ways and don’t see them improving any time soon.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.