This question has come up before on the old forum, and has always been a bit of a fudge- The BS only specifies the breaking times max and min at a few spot currents, and allows makers to thread their curves between these in anyway they see fit.  This makes the spec as loose as a goose, and to guarantee discrimination is not really possible.

The other fixed requirements are the dimensions,  the safe breaking capacity of 6000A, and a self heating upper limit of 1 watt at full load.

The makers curves are useful however, as they give us an I2t value, or at least it is hiding in plain sight.

At the fast end of the curves, pick a point, and then verify that at half that current the time to break is 4 times longer - here we are in constant I2t territory.


On the no-break curve we get (5* 13A) ^2 * 0.01seconds  = 42 A2 seconds for 13A fuse

On the all-break curve we get (20*13)^2 * 0.02secs = 1352 Amp squared seconds.   (see  that is a touch over  a 30 to 1 range  - you will  see what I meant about a loose spec,)

We cannot assume any energy limiting action, as  it is not specified, but it is unlikely to be great  even at the max permitted 6kA fault level anyway..

We can use the 1 watt at 13A to deduce a resistance of a cold unblown fuse to be  rather less than 1/(13^2) or about 5 milliohms. For good practical reasons  of length & cross section of fuse wire it is always more than half this, perhaps 3 milliohms . This is small compared to the loop impedances of 35milliohms or so required to give  6kA fault current from 230v, and can safely be neglected in most cases. The 6kA limit is met even on an ideal supply, if there is more than a metre of so of 1mm2 cable in the way, or 2.5metres of  of 2.5mm2 cable etc.

I would partly agree with this (in disagreeing with the statement "We cannot assume any energy limiting action") as the way fuses work is inherently energy-limiting. It takes a certain amount of energy to blow the fuse and energy in excess of this will not be let through. However if it was slightly reworded as "we cannot assume a known energy limiting action" I would agree with it.

My only problem with Mike's method of calculation is I suspect it falls down at very high currents. Because the fusible link inside the fuse melts, at extremely high currents it will let more energy through as the arc is extinguished than at lower currents. However the time current calculation breaks down when you get below the duration of a half cycle (10ms) as the current doesn't (or at least may not) reach the peak prospective short circuit current as it is interrupted while still rising.

Alasdair

Sorry - I agree I was a bit fast and loose with that one.


What I was trying to say, and didn't make at all cleat,  is that there is no equivalent to a curve like this  one


For the BS1362. So we do not know where along the line of increasing PSSC to assume a departure from a constant I squared t.

The only safe thing is to assume a constant energy (I2t = constant) and no roll off at high fault levels, and no effective reduction in PSSC down stream of it. Though I agree it is not true.