MIT Extracts Power From Temperature Fluctuations

As a civilization, we are proficient with the “boil water, make steam” method of turning various heat sources into power we feed our infrastructure. Away from that, we can use solar panels. But what if direct sunlight is not available either? A team at MIT demonstrated how to extract power from daily temperature swings.

Running on temperature difference between day and night is arguably a very indirect form of solar energy. It could work in shaded areas where solar panels would not. But lacking a time machine, or an equally improbable portal to the other side of the planet, how did they bring thermal gradient between day and night together?

This team called their invention a “thermal resonator”: an assembly of materials tuned to work over a specific range of time and temperature. When successful, the device output temperature is out-of-phase with its input: cold in one section while the other is hot, and vice versa. Energy can then be harvested from the temperature differential via “conventional thermoelectrics”.

Power output of the initial prototype is modest. Given a 10 degree Celsius daily swing in temperature, it could produce 1.3 milliwatt at maximum potential of 350 millivolt. While the Hackaday coin-cell challenge participants and other pioneers of low-power electronics could probably do something interesting, the rest of us will have to wait for thermal resonator designs to evolve and improve on its way out of the lab.

[via Engadget]

51 thoughts on “MIT Extracts Power From Temperature Fluctuations

      1. There have been extensive research in Low Temperature Differential Stirling engines, but they are also big and slow, for the power they produce, but still better than Thermoelectrics! I think the reason they used them here was to have something easy measurable for the experiment. Maybe it would be better to use an organic Rancine cycle, like a heat pump with ground loops run in reverse?

    1. According to the linked article: “This perpetual difference between the two sides can then be harvested through conventional thermoelectrics.” Sounds to me like it _is_ a peltier. Looks like there “secret sauce” is all about managing the material properties to create a long-term thermal gradient for the peltier to exploit. Effectively, last night’s “cold” against today’s “hot”.

      1. There’s a bunch of phase changing wax that acts as a thermal mass in an insulated box, so the heat can only enter and exit through the peltier elements. Since there’s this thermal resistance along the way, there’s always this lag no matter what you put in the box – the wax is used simply because of its high specific heat capacity.

        1. Oh! A thermal battery! Those are so cool, and arguably much more interesting than the little peltiers that they’re attached to. They actually store MUCH more thermal energy than simply their specific heat capacity would be able to. They pick a material that melts within the desired working temperature, and then it takes/dumps a TON of energy to cause the phase change, way more than just a regular old thermal mass. Very cool!

          The MIT article also says this is the world’s first “thermal resonator.” It isn’t. It just isn’t. Stuff like this has been done lots, they’re not breaking new ground here. I’ve seen at least two other designs that are nearly identical in the past. Thunderf00t even did a video about one of them last year. I saw a completely passive design a while back (zero electronicals) that was also based on the day-night cycle, but was supposed to collect water. I don’t know of the thermodynamics were sound, but it was extremely ingenious in the way they were accomplishing it.

          1. The phase changing material in this case doesn’t have a definite melting point, because they can’t constrain the temperature swing to occur over that particular range. The other problem of such materials is the hysteresis or supercooling effect, where the material doesn’t start to freeze/melt until the temperature has passed several degrees, sometimes tens of degrees C past the actual melting point, so using the stuff over a small temperature swing is impossible.

            So instead they’re using a material which undergoes a gradual transformation over a range of temperatures.

          2. Yea, the most suspicious expression seems to be “thermal resonator”. Resonance is periodically “nudged” oscillation. In thermodynamics heat flows from hot to cold with some resistance and direction but there is no “thermal momentum” so how could such a system possibly oscillate on its own and consequently resonate. Sounds like snake oil to me. I’d like to be wrong, but every instinct tells me I’m not. :-)

    2. It is Seebeck effect (Peltier is to motor as Seebeck is to generator); It’s just not as picky about specific inputs and can work over a range, as long as there IS a range to work in.

      1. Correct, but they are likely using components commonly sold as “Peltier devices”. As far as I know, there’s no difference in the device itself, just two different names for the effect, depending on direction.

      1. They are my favorite clocks in the world. I am one of the lucky few in North America that can buy parts for them direct from the factory, and am authorized to service them. At last count I think there are 2 of us in Canada that can get the parts.

  1. If one might be conducting a covert activity in a residence or motel tying onto the hot/cold water lines may produce enough
    power to charge a battery used intermittently for such activities.

  2. It seems to me you would have a much larger delta T just placing the Peltier devices in a wall on the side of a house. Inside=70F outside =20F. (Winter) which gives a 50 degree difference and something to work with to get useful energy out of these things. I want to install and entire wall on the side of my house to see how much energy I can get day and night in winter, summer, fall and spring.. I might try this in my next house.

    1. In that case you are harvesting energy from your heater by passing heat out through your wall. This is a great way to reduce the R-value of your house and spend more energy on heating or cooling.

      1. The Peltier works from multiple thermal couples and having a larger difference in temp. between the two metals yields more power generated. There might be minimal losses through heat exchange but otherwise I would be harvesting energy normally lost through the roof. The aluminium oxide covering the thermal couples is a very good insulator. I have several of these units, I might experiment to see what the heat exchange losses might be.

        1. Still just harvesting energy from the heater. Thermodynamics doesn’t give out freebies. It would be better to upgrade your insulation, turn down your heater, and just use the electricity correctly the first time. I’ll bet insulation is cheaper than a bunch of peltier equipment.

          1. How about using the OUTSIDE brick or concrete wall on the sunny side ie north in Australia and then putting hundreds of these devices against it and then sandwich them with a black plate of aluminium. That way you are “comparing” the difference of the nights cool against the present sun/wind temps and vice versa at night as the heat will eventually transfer through to the wall just in time for the outside air temps to fall and the wall to heat?

    2. The peltier device leaks about 20 times more heat through than it returns in electricity – it generates electricity from the flow of heat – so you’re really just running a really expensive generator on whatever fuel it is you’re heating your home with.

        1. The peltier leaks more thermal energy through it than a empty piece of drywall and fiberglass.
          To keep the a windowless drywall-and-fiberglass house at temperature X will require less energy than one where all the walls are peltier junctions.
          The power you get out of a peltier junction has to come from somewhere, and that somewhere is probably the thermal difference that you already made using a furnace or air conditioner.

          If you’re talking about adding peltier junctions in thermal series with your existing insulation, you’ll instead find that you can’t maintain a temperature difference across the peltier junction. Thermal resistances in series divide temperatures the exact same way ohmic resistances divide voltages, and the peltier junction has a tremendously lower thermal resistance that the fiberglass.

        2. Think of it as a hydroelectric dam that is being pumped up from a lake below. You gain head height from the energy you spend pumping the water uphill, but you can’t get energy out from just the difference in water levels, just like you can’t get energy out from just the temperature difference between the interior and exterior of your house.

          In the case of the dam, you get energy by letting the water flow back downhill, which lowers the water level in the reservoir. In the case of the peltier element, you get energy by letting the heat flow out of your house through the device, which lowers the temperature indoors.

          You spent so much energy heating up your house to a comfortable temperature, and if you poke a hole in your insulation by installing the peltier element, you’ll need to turn up the heater to keep the temperature up, and so you’re wasting energy to the inefficiency of the peltier element.

        1. Yes, a small thermoelectric generator would let only a small amount of heat energy be lost, but it would produce even less electrical energy. Maybe it would make better sense if I explain it using one of my sillier “inventions” as a kid. I thought that a way to make “free” electricity would be to attach a small generator to one of the wheels of a car, where the motor would spin when the wheel spins. I reasoned that the loss of energy from the vehicle was minimal, so it would cost virtually nothing. However, for every extra Joule of mechanical energy the car engine would have to output, my generator would only be able to extract a fraction of that.

          This all part of the 2nd law of thermodynamics, where a closed system (the universe) tends towards increased chaos. This means that for every energy transformation (for instance the conversion of the energy stored in the temperature difference across your wall to electrical energy via a thermoelectric generator), some amount of that useful energy would be lost, or “degraded”. Your system would cost very little to operate, but it could never recoup the energy costs.

          That being said, your idea isn’t entirely flawed. The introduction of more efficient thermoelectrical generators could allow for energy to be recaptured more easily. Instead of capturing energy from already useful processes (like the moving car or the warmth of your house), we could capture energy from things we don’t want to be hot. For instance, you could potentially extract useful energy from the temperature difference between the exhaust gasses of a furnace and the surrounding cold without worrying about that energy extraction detracting from the actual purpose of using the furnace: heating your house.

          1. You don’t want to cool the flue gasses of your furnace, because they won’t rise up and first you end up with water condensing inside your chimney, and then you end up with the neighborhood being blanketed by the flue gasses.

            It can even flow back down into the house and fill it up with CO2.

        2. Heat exchange is what generates the electricity!
          If you wanted your home to stay at the same temperature you would have to supply more heat. At the minimum ten times more energy than you would harvest. Then there is the cost of the Peltier device itself, 30 year payback.

    3. I’m tempted to bolt a Peltier cooler to my chimney. In winter the heat of burning wood contrasts with the cool outside air, and in summer the cool chimney brick contrasts with the warm air. It wouldn’t be a lot, but maybe enough to contribute to landscape lighting.

  3. I think of the project where they melt sand with mirrors and then use the melted sand to make steam to run a turbine would be more efficient. IIRC, they can melt enough sand during the day to continue making steam through the night.

    1. A device like this could work for longer than solar, TNG and Stirling engine in space missions. Drop one on Mars and if it only collects a few Wh/day, it would still be sufficient to send telemetry to Earth periodically, but should work for centuries.

      1. Agreed – the greatest promise here isn’t about cost effectiveness or efficiency, but its ability provide long lasting power for very specific niches. Any place with a regular hot-cold cycle becomes a candidate, as a thermal resonator can be designed to take advantage of the specific temperature range and time period.

        Mars is one possibility, but we can go further. Want to drop a lander on Europa? Between risk of RTG contamination and mechanical complexity of deploying a huge solar array, a thermal resonator might be a viable power source.

  4. As usual Mit news paper is useless, no details at all . It is some kind of advertisement to be read by managers , lawyers or strategy advisers. People with no scientific or technical formation but who runs the money !

  5. This is probably a thesis experiment. My question is, who approved this as a thesis experiment? There’s nothing new here… not even putting the thing out in the yard, or more likely, on the roof.

Leave a Reply to rnjacobsCancel 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.