Fig 12b. page 26. Only just notice that they say they are "inconel" resistance wires. But how can the resistor wire, external to the reactor, glow more brightly than the reactor itself (implying it is hotter)? Also if the reactor external surface is at 1250-1410°C, then on top of the 30-50°C temperature drop through the walls the resistive wire would have to be even hotter (up towards 1500°C) in order to be able to radiate the 900W! And there are no "inconel" metal wires that can survive such temperatures in air for a month. http://www.engineeringtoolbox.com/melting-temperature-metals-d_860.html
There is also a mistake in their discussion for these figures, the metal resistors within the reactor would necessarily have to be hotter than the reactor itself if they are to be able to output heat into the reactor. But then again at >1450°C inconel resistance wires would have turned into puddles anyway. The 2mm high fins on the outside, subjected to convective air cooling would be 30-50°C colder than the base of the fins due to the longer conduction path and hence greater temperature drop through the alumina. And that should influence the thermography to under-read, but also means the temperature within the reactor would have to be even higher to create 1400°C average surface temperature. So this proves that the thermographic temperature readings are off by a very long way - at least 150-200°C too high if inconel wires that max-out at about 1300-1350°C were to have survived for a month. We now know there is a huge error, we just don't know the source of the error, or how big the error is. But eg 1210°C vs 1410°C would drop power output by 40% from that claimed so that the COP would be <2. With the existence of such a huge error of unknown source the claims of power output cannot not be trusted at all. On the upside, it does however mean that the nickel was not melted, and lithium vapour pressure was a lot lower.
