The current carrying capacity of a cable refers to it carrying a continuous load.
An earth cable normally carries no load, and under fault conditions will carry a significant instantaneouscurrent but only for a
short time most Regulations define 0.1 to 5 sec before the fuse or breaker trips. Its size therefore is defined by different calculating
parameters.
The magnitude of earth fault current depends on:
(a) the external earth loop impedance of the installation (i.e. beyond the supply terminals)
(b) the impedance of the active conductor in fault
(c) the impedance of the earth cable.
i.e. Fault current = voltage / a + b + c
Now when the active conductor (b) is small, its impedance is much more than (a), so the earth (c) cable is sized to match. As the
active conductor gets bigger, its impedance drops significantly below that of the external earth loop impedance (a); when It is quite large its
impedance can be ignored. At this point there is no merit in increasing the earth cable size
i.e. Fault current = voltage / a + c
(c) is also very small so the fault current peaks out.
The neutral conductor is a separate issue. It is defined as an active conductor and therefore must be sized for continuous full load. In
a 3-phase system,
If balanced, no neutral current flows. It used to be common practice to install reduced neutral supplies, and cables are available with
say half-size neutrals (remember a neutral is always necessary to provide single phase voltages). However the increasing use of non-linear
loads which produce harmonics has made this practice dangerous, so for example the current in some standard require full size neutrals.
Indeed, in big UPS installations I install double neutrals and earths for this reason.