The dream of the ancient alchemists may come true as Marathon Fusion announces that its tokamak fusion reactor technology can turn common mercury into gold as a byproduct of fusion operations in quantities that would make Auric Goldfinger blush.

  • qupada
    link
    fedilink
    124 days ago

    The units are weird, and the person writing the article seems to have conflated a few different quantities.

    From the actual press release linked in the article:

    Using our approach, power plants can generate five thousand kilograms of gold per year, per gigawatt of electricity generation (~2.5 GWth), without any compromise to fuel self-sufficiency or power output.

    So unless I’ve also missed something, what they actually mean is 5 tons per year assuming a continuous power output of 2.5GW, which is roughly 22TWh of energy generation.

    Or in slightly more approachable units, approximately 0.23g/MWh.

    • @[email protected]
      link
      fedilink
      English
      24 days ago

      GWth means 1GW of thermal energy, nothing to do with tons.

      The paragraph of note from the preprint paper:

      Under this simplification, there are ∼ 1.12 × 1028 fusion reactions per year in a 1 GWth fusion device. Assuming neutron multiplication is dominated by (n, 2n) reactions, in order to achieve a TBR = 1.2, at an absolute minimum 20% of all fusion reactions must have a corresponding (n, 2n) reaction in the blanket; as a less conservative value, it is known that simplified blanket configurations (2 m thick, no structure, natural Li) can achieve as high as TBR ∼ 1.85 [26], in which case at least 85% of fusion reactions must have a corresponding multiplication reaction. This range corresponds to 3.7 × 103–1.6 × 104 mol/yr, or for a product with a mass of 197 amu, 732 − 3114 kg/yr of material production.

      The paper seems to report an upper bound of 3000 kg/GWth/yr.

      There does seems to be some conflating of GWth (GW of thermal power) with GWe(GW of electrical power). Assuming an efficiency of ~60% would make the numbers line up and that seems in the ballpark of possible conversion efficiencies.