BS 1362 ("plug top") fuses.

The two curves represent  the 'all fire' and the 'never fire' conditions - real fuses have a considerable spread due to wire tolerances and variation in alloy mixture and so forth, and then on top of that the environment, i.e. how well the fuse caps are cooled, has an effect at on the exact boundary shape - if the fuse is running hot, then it  takes less of an overload (time or duration) to melt it.

Actually all fuses and breakers are like this really, just for BS 7671 we are more interested in the 'all fire' condition, and concentrate on that curve, and often assume that the other limit might as well be a single value, the rating on the fuse body.

However the let through energy you need to get good discrimination between final circuits and distribution sort of relies on the same idea - just expressed differently.

Oh, and at 7 A there will be selectivity - the 6A breaker wont, the 3A Fuse will, after any duration that could be up to about 20 mins... I agree for a (near) short circuit the breaker will win up to about 500A, the the fuse will take over again as the breaker mechanics cannot go supersonic... . But do we really need to care that much?  because regardless of which operates, the cable being protected will be fine.

Edit.

Once upon a time we just assume double or treble -  so the 'never fire' curve for a 30A BS3036 hot wire fuse, may be assumed to be somewhere between 30A and 60A forever, and 60-120A for some minutes. etc.

And before anyone winces - 13A fuses discriminate really well with 20 and 30A fuse wire, and that in turn discriminates really well against a 60A or higher company fuse - and for the best part of half a century that was the preferred line up, and in many buildings in daily use,  it still is. There is a reason there is so much slack in the cable ratings, and kettle leads can be 0.75mm without over heating, when according the annex they are way undersize.

Mike.

I think the basic answer is because it's a direct duplicate from BS 1362. For most purposes in BS 7671 (typically disconnection time for shock protection, and adiabatic), we are concerned with the right-most curve for each area.

Thank you Mike and Graham for your very helpful and prompt replies. Regards, Colin.

I agree that a 3 amp fuse and a 6 amp MCB wont reliably discriminate.

I would still fit the 3 amp fuse, for leagal, liability and warranty reasons. If the instructions state "fit a 3 amp fuse" I would do exactly that.


In the probably rare event of a fault, having to check the fuse and the MCB is no big deal.


"Fit a 3 amp fuse" is a nice simple instruction for the plumber or gas fitter who installs the boiler. Avoids any doubt or ambiguity.

A more complex instruction such as "fit a 3 amp fuse unless the supply is from a 6 amp MCB" Increases the risk of mistakes.

Don't manufacturers' data for MCBs have similar zones? I must admit that I find them confusing.


Putting two and two together, an MCB will not trip in under an hour if I < 1.15 x I_n but we should not take advantage of this (433.1). The situation might be unmanageable if an MCB tripped the moment that it's rated current was exceeded.


Equally, an MCB must trip within an hour when I > 1.45 x I_n assuming that the lowest CCC of any conductor equals the rating of the MCB (433.1.1(iii) ).


In reality, any particular MCB will trip somewhere within the zone. Precisely where cannot be said because of manufacturing tolerances.


If I have misunderstood this, please let me know.

Don't manufacturers' data for MCBs have similar zones?

They do indeed - e.g. https://library.e.abb.com/public/114371fcc8e0456096db42d614bead67/2CDC400002D0201_view.pdf

  - Andy.

Actually it not just an accident, it is essential for sanity that MCBs and fuses are not too trippy and have a well defined load dependant time delay -  many domestic devices have significant inrush currents that persist while motors are spinning up (Fridges may take 50-100A until the rotor has got moving) or while hot wires are getting hot.The old filament lamp is a classic for that temperature dependant resistance - a 60W lamp takes about 1/4 of an amp when hot, but has about 10% of the resistance when cold. so takes more like 2 and a bit amps when cold - so dimmers on the really dim setting are always  dealing with thin slices of many times the current you might expect, so you find you need a 10A triac in a 200W dimmer...


Modern LED  lights with  switch mode supplies introduce a new issue, of a bridge rectifier feeding a capacitor, where on switch on (mains - 2 lots of 0.7V ) finds itself charging a flat capacitor of many microfarads, so you may get 100A plus inrush for the first part cycle.

The breaker curves are deliberately designed to  approximate a fuse-like time current integral up to the point the magnetic near instant part takes over to allow things like motors, lamps etc to be used without too much hassle.

Mike

Thank you to all for the helpful and thoughtful replies; I understand exactly now why there are two curves published for the fuse. I always refer to the time/current charts in BS 7671 and hadn't thought too much beyond these for quite a while. Much appreciated. Regards, Colin.