Gas Cooled Fast Reactor: The GFR (Fig. 4) is a variant of the SFR system described above, but employing a gaseous coolant—helium usually being proposed, although CO2 and steam have also been suggested.

The selection of a gaseous coolant avoids some of the difficulties of a two-phase fluid, most notably the sudden increase in reactivity that accompanies coolant boiling (known as the void coefficient).

If helium is adopted, the difficulties associated with a chemically reactive coolant are also avoided, as are problems of corrosion and stress-corrosion.

In addition, the absorption cross-section of helium is lower than liquid metals, providing a harder (higher energy) neutron spectrum, especially compared with sodium.

Finally, higher coolant temperatures are possible compared with liquid metals, and a direct-cycle gas-turbine arrangement is possible, offering high generating efficiency or the possibility of high-temperature process heat.

Unfortunately, the benefits of the GFR also provide some of the greatest technology challenges. Gases offer relatively poor heat transfer properties.

In thermal reactors this difficulty is mitigated by the massive graphite structures which provide a large thermal inertia, limiting the heating rates during transient events such as a
loss of coolant.

In a fast reactor, where a compact core free from moderating elements is vital, the poor heat transfer characteristics of a gaseous coolant place very severe requirements on the ability of the fuel and structural components to survive extremes of temperature.

The GFR can benefit from the gas-cooled technologies established as part of the HTR and UK gas-cooled reactor programmes, and from elements of the SFR fuel technology which
has also already been established.

However, it is clear from the above that the fuel developed for the SFR (or other metal-cooled fast reactors) will not be adequate for the GFR.

In particular, it seems difficult to imagine that metallic cladding will be adequate for GFRs, or that the decay heat rejection systems from SFRs will be appropriate for GFRs.

Finally, it must be noted that unlike the VHTR, SFR, LFR, and MSR, no experimental or prototype GFR has ever been built.