Defense Intelligence Reference Document Inertial Electrostatic Confinement Fusion

UNCLASSIFIED/;SflHl 8FFIIItliL '1181!! 8HLV
results. Rider was particularly interested in the claim that due to its beam-like non-
Maxwellian plasma, the IEC can burn "advanced fusion fuels" such as D- 3 He and p- 11 B
easier than traditional Maxwellian type plasma devices (Tokomaks, etc.). Thus, he
considered use of D-T, D-D, D- 3 He, 3He-3 He, p- 11 B and p- 6Li fuels. Due to their high Z
components, all of these fuels must battle large energy losses via Bremsstrahlung.
These losses were evaluated using the traditional formula, but his evaluation has a built
in bias since the loses depend heavily on the electron ion temperature ratio which in
turn depends strongly on the values assumed as already discussed. Deviation
from an equilibrium electron energy distribution also strongly affects radiation emission.
Using the Maxwellian average values, he found that Bremsstrahlung losses would be
prohibitively large for 3He- 3He, p- 11 B, and p- 6Li reactors and will be a considerable
fraction of the fusion power for D- 3He and D-D reactors limiting use to D-T. As a
corollary, he concludes in contradiction with earlier claims that it does not appear
possible for the dense central region of a reactor-grade IEC device to maintain a
significantly non-Maxwellian ion distribution or keep a low electron to ion temperature
ratio. The problem, however, is that the assumed rate constants would naturally force
this conclusion. Further, these rate constants lead to Rider's build-in result forcing the
ions to form a Maxwellian distribution with a mean energy close to the energy of the
potential. Consequently, in his analysis, ions in the energetic tail of the distribution are
lost at rates faster the fusion rate, giving low Q values.
Rider considered the Polywell type IEC and even with exceedingly optimistic
assumptions about the potential well, he found the electron losses are intolerable for all
fuels "except perhaps DT". Based on these results, Rider concludes that for the IEC
system to be used as a fusion reactor it will be necessary to find methods to
"circumvent these problems, especially the excessive Bremsstrahlung losses". Certainly
reducing radiation losses should be an ongoing study, but his pessimism appears to be
overdone.
The problem with Rider's analysis is his not using reaction and scattering rates
averaged over the non-Maxwellian distribution characteristic of an IEC reactor. This
includes both the beam-like ion distribution and the large ion-electron temperature
ratio. This very basic energy analysis should be redone with revised reaction rate data,
but to date have not been reported. Despite questions about Rider's analysis and
pessimistic conclusion, his recommendations of issues to study and overcome remain
quite valid.
NEUTRON SOURCE SIMULATIONS
Several simulation studies have focused on neutron source type IECs. In this case, as
opposed to future power reactors, the background gas is of sufficient pressure,
resulting in beam-background scaling of the fusion rate (i.e., theoretically this gives a
current x pressure scaling). Charge exchange also becomes a significant factor in device
performance. In Reference 4.8, Miley et al. used an analytical model of charge-
exchange collisions in the IEC plasma to include ion time-of-flight and fusion neutron
generation rates. Results from the model simulating 10 mA of o+ ion current in a 30-cm
diameter IEC device at 50 kV matched the experimental results of 106 fusion neutrons
per second. The model was also used to find the effects of grid diameter on neutron
yield and show that the yield scales as grid diameter is raised to the power -0.41. This
factor is very close to the experimental scaling observed from the UIUC neutron source
44
UNCLASSIFIED,' }P8"1 8PPll!l"'I! l!l!il! 8111!¥