High Temp Heat Engine Achieves 40% Efficiency

People generate lots of waste heat. It makes sense that there is a desire to convert that heat into usable energy. The problem is one of efficiency. Researchers from MIT and the National Renewable Energy Lab have announced a new heat converter that they claim has 40% efficiency. Of course, there’s a catch. The temperature range for the devices starts at 1,900 °C .

The thermophotovoltaic cells are tandem devices with two cells mated on one substrate. Each cell is multiple layers of very thin and somewhat exotic materials. So this probably isn’t something you will cobble up in your basement anytime soon unless you’re already manufacturing ICs down there. It appears that the secret is in the multiple layers including a reflective one that sends any missed photons back through the stack.

The paper is pretty dense, but there’s a Sunday-supplement summary over on the MIT site. Using heat storage leads to the ability to make heat batteries, more or less, and harness what would otherwise be waste energy.

We’ve noticed a lot of interest in drawing power from hot pipes lately. All of them techniques we’ve seen rely on some kind of exotic materials.

17 thoughts on “High Temp Heat Engine Achieves 40% Efficiency

  1. There’s an entire field of waste heat thermoelectric energy harvesting. Reference that as an intro is unfortunate, if not bait-and-switch.

    Anyway, a tangible comparison can be made between TPV cells and an out-of-band light recycling structure intended as a light source:

    Finally, one could consider these structures related to halogen reflector lamps, except those are designed to only reflect visible light and let IR escape at the back.

  2. So it is a stack of IR solar cells each tuned to slightly different IR frequencies with a gold mirror at the bottom (gold reflects IR really well 97-99% but melts at 1064.18 °C ; ​1947.52 °F). I wonder could there be a cheaper version of this with aluminium (aluminium reflects about 95-99% but melts at 660.32 °C ; ​1220.58 °F) or silver (reflects about 96-99% but melts at 961.78 °C ; ​1763.2 °F). Maybe add a high temperature TEG that is transparent to IR sitting above the aluminium/silver/gold mirror to allow for cooling (it would reduce the efficiency slightly, but extend the useful lifetime). Now I find myself wondering how well does liquid gold reflect IR.

    (ref: https://en.wikipedia.org/wiki/File:Image-Metal-reflectance.png )

    1. The gold is the electrical connection, not the reflector. The reflector is the interface between the AlGaInAs layer and the GaInP layer due to their difference in refractive index. It is a dielectric mirror (Bragg mirror).

      Aluminum ions migrate at high temperatures in Gallium based semi-conductors, so it would not be usable in this particular application.

      1. “A gold mirror on the back of the cell reflects approximately 93% of the below bandgap photons, allowing this energy to be recycled.”

        I’m thinking this is the broadband case for all wavelengths below the band for which one could optimize a distributed Bragg reflector.

        “In all cases, the cell temperature is 25 °C.”

        Can we please stop talking about diffusion and molten back metallisations now, please.

      1. Even with 97% reflection, these things will warm up. They may start out at room temperature, but they will slowly heat up without cooling (and at higher temperature they typically become less efficient).

        The interesting part for me is that this is ~40% efficient. The best steam turbine has a gross efficiency of ~49 percent (ref: https://www.power-eng.com/news/new-benchmarks-for-steam-turbine-efficiency/ ).

        So use solar concentrators to heat a working fluid to a high temperature, store the fluid once at temperature within several meters (yards) of insulation and then use TPV cells to generate electricity when needed. It maybe less efficient than using direct generation, but the ability to offset generation until needed and the ability to instantly ramp up or down generation definitely makes it an interesting technology.

        1. The difficulty is, it takes a whole lot of energy to get anything 1,900 C hot and it only starts to store energy efficiently above that temperature. Then the next problem is, what materials to use. Molten salt storage isn’t an option because they boil off well before 1,900 C.

          1. @Dude
            As the working fluid maybe Tin/Lead/Bismuth:
            Tin would be good (Melting point 505.08 K ; 231.93 °C ; ​449.47 °F, Boiling point 2875 K ​; 2602 °C ; ​4716 °F)
            Bismuth (Melting point 544.7 K ; 271.5 °C ; ​520.7 °F, Boiling point 1837 K ; 1564 °C ; ​2847 °F)
            Lead (Melting point 600.61 K ; 327.46 °C ; ​621.43 °F,. Boiling point 2022 K ; 1749 °C ; ​3180 °F)
            You loose out on the typically high phase change energy of molten salts, which is really efficient in terms of energy storage.

            Lead Heat of fusion 4.77 kJ/mol
            Tin Heat of fusion white, β: 7.03 kJ/mol
            Bismuth Heat of fusion 11.30 kJ/mol

            There is nothing stopping you from using a very high pressure gas either, the system could be liquid/gas and not solid/liquid or solid/liquid/gas for more phase change energy storage.

            And possibly tungsten/titanium/platinum pipes and storage tanks would be a lot of fun. The security required would be interesting.

          2. The idea is to use solid carbon heated in a sealed, insulated vessel, which then radiates the IR. Carbon is actually good at radiating; the reason kerosene lanterns can be so bright is the glowing carbon articles in the flame.

  3. Note the advertised performance *starts* at 1900 C. Operating range is optimized for 1900-2400 C.

    Not many fuel burners can efficiently create such temperatures, at least not when burning in ambient air.

    One standout: Plutonium oxide stays solid well above these temperatures. This thermophotovoltaic converter would make a fantastic radioisotope thermal generator, if they could get reasonable lifetime out of the material.

  4. Oh FFS. This is not a heat engine, as it does not create any mechanical energy. It is just a photovoltaic cell. And 1900C is not a high temperature, as PVs go, it is a low temperature – most PVs operate below 700nm, which corresponds to a radiation temperature thousands of degrees hotter.

  5. I think lots of you guys are mistaking the operating temperature of the PhotoVoltaic cell (which I guess can be even room temperature) with the black body temperature of the radiation source (those 1900 degrees Celsius).

  6. Just thought I’d also mention that it clearly mentions in the abstract that the *emitter* temperature is what they’re referring to, not the cell.

    “The TPV cells are two-junction devices comprising III–V materials with bandgaps between 1.0 and 1.4 eV that are optimized for emitter temperatures of 1,900–2,400 °C”

    Wouldn’t make sense for the device itself to operate at those temperatures, since most things in that stack melt well before those temperatures. A little weird for the temperature to be listed in C instead of K, but whatever I guess.

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.