yes, instantaneous conversions are confusing... you have to mix energy (joules sitting in the capacitors pre-pulse) with power (watts flowing into the Polaris system to do various pulse-related things such as compress or accelerate the plasma (essentially transferring the energy), then flowing back out into the capacitors to become joules again)
typically capacitance seems to be measured in terms of the length of time they can reliably deliver a certain amperage at a specific voltage, but Helion's setup is obviously not a typical use case... they gave us capacitance in joules which probably makes more sense in this context (where we ultimately care a lot more about total energy in vs out)
although I think my J/W conversion in the prior post was for 1 second of wattage which is certainly way longer than the pulse, so maybe up those by a factor of 100 - 1000
we don't know (or at least I don't) the precise planned length and wattage of the Polaris pulse, except that it must be less than 100ms since their planned frequency is 10/sec, so my guess was a plasma lifetime of about 1 ms, so if they used the whole bank of 50MJ over .001 seconds we'd get 50GW for the pulse cost I think, and 250GW as the max resulting power
that's a hot tamale!
fwiw Tri-Alpha claims to have gotten FRC plasma lifetimes as long as 1ms, although obviously their setup is quite different
https://nucleus.iaea.org/sites/fusionpo ... nt0644.pdf
there are apparently some schemes for recapturing brem (and I can't rule out significant recapture of the compression energy by intrinsic (non-fusion-related) plasma pushback) but to Kirtley's point, Helion isn't relying on fusion outputs to continuously heat the plasma so losses have a very different context
of course for Polaris the big question seems to be whether they can reach >20KeV plasma temps, where Q is plentiful by any definition of losses
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...