Well, I've already done some real load tests with cords of my own (a few of which I've described
in a previous thread on this forum). (I probably ought to assemble a cleverer arrangement than convection heaters for future tests, though, in order to maintain
slightly lower room temperatures during Australian summers
.)
So, on the basis of those: For cords up to about 10m long, then, our 10A through 1.0mm^2 appears reasonably safe (although up to 25m is permitted, which will be somewhat worse). But
16A through 1.0mm^2 for
20m sounds like a meltdown waiting to happen, if the user isn't thinking.
On the occasion when you need a still-longer cord, then, I would go with a
thermally protected extension reel, and/or upsize the wires another step (e.g. 1.5mm^2 for 10A, 2.5mm^2 for 15/16A).
While laid straight, though, the flexes themselves seem to cope well enough with 13A through 0.75mm^2, or 16A through 1.0mm^2 (or 20A through 1.5mm^2 etc.).
I
strongly suspect that (most of) the true problem behind the VDE's "upgrade" was related to the inconsistent quality of terminations (which is no doubt severely exacerbated by penny-pinching). But although thicker wires may well keep borderline terminations a
bit cooler and survive for a little longer, I've clearly seen signs of localized overheating even inside appliances that do it 'by-the-book' with 1.0mm^2 for 10A (or even lower current) -- and
usually while the appliance itself still works.
(For obvious reasons, space heaters are among the foremost victims...)