Whoever quoted that has taken it out of context ! -

there is a formula and a table that really says the bonding size  can be a submultiple of the live conductor size, but this rule then only should be applied between upper and lower limits = too small and the wire breaks - so there is a minium size, too large, and presumably the street-main is only that  large size due to a long run and volt drops, not current rating and thinner bonding is OK, as other factors limit the fault current.

Mike.

Thanks for your reply; I think it may be my interpretation of "should not surpass" !

Table 54.8 describes CSA requirements as "Minimum copper equivalent cross-sectional area of the main protective bonding conductor".  

The maximum value in the table is 50 mm² but I don't see anything that says that this can't be more than 50 mm².

It's many years since I looked at any of this to be honest and it's abstract anyway.

But when I read the article, it just seemed odd to have a notional maximum copper conductor size since the only obvious rationale for doing so would be to limit current and there's no good reason I know of to do this when setting an equipotential zone.  

Thanks for your reply; I think it may be my interpretation of "should not surpass" !

Table 54.8 describes CSA requirements as "Minimum copper equivalent cross-sectional area of the main protective bonding conductor".  

The maximum value in the table is 50 mm² but I don't see anything that says that this can't be more than 50 mm².

It's many years since I looked at any of this to be honest and it's abstract anyway.

But when I read the article, it just seemed odd to have a notional maximum copper conductor size since the only obvious rationale for doing so would be to limit current and there's no good reason I know of to do this when setting an equipotential zone.  

And there is nothing to stop you using a larger conductor if you wish to do so - it is more that the recommended size stops going up at 50mm2 ,because  substations and fault current stop getting bigger much beyond the 1MVA/ 30kA level, so the rule of thumb has a limit.

Also of course, unspoken, is an assumption about the length of that bond, and the area of the equipotential zone there may be corner cases where it is not the right thing to do.

Mike.