Why does fuse blow up sometimes?

IMPORTANT:

• It is far far better that a fuse sometimes blows when there was no need for it to, than for it to sometimes fail to blow when a fault condition exists.

Blowing a fuse. The term "blow" will be used here for the fusing of a fuse - the act of melting the fuse wire and breaking the electrical circuit. Terms such as "it blew a fuse" and "why did the fuse blow?" are common here. The term "blow" in this context may be less common in some countries. Using "fuse" which is correct, as in "the fuse fused", is liable to be too confusing :-).

Why do they blow?
Should they?
The purpose of a fuse is to protect equipment and wiring against the damaging effects of electrical faults which cause excess currents, and to disable equipment which is faulty. The fuse "blows" when the current carried exceeds the rated value for an excessive time. The higher the overload the shorter the period before the fuse blows. So, equipment which is meant to "draw" 10 amps but which has a short from phase to ground, so it draws, say, 100 amps, will blow its fuse in milliseconds. But, a piece of equipment which draws say double the fuse's rated value, may take many seconds to melt the fusewire and to blow the fuse. The ratio between trip times(time to blow) and "overload to rated current ratio" vary with fuse design and can to some extent be controlled by the manufacturer. This is a complete subject in its own right, but assume that a fuse will blow "after a while" at 2 x + overload and will blow almost immediately with say 10 x + overload.

A piece of wire can only be so smart ...
Because a somewhat complex task is being carried out by a deceptively simple piece of equipment (ie a piece of wire) and because the fuse is not always optimally dimensioned for the equipment used, the fuse sometimes "blows" when there is no significant or long term fault condition present.

To blow or not to blow ? - that is the question.
Dimensioning & surges.
Assume that a fuse will blow "after a while" at 2 x its rated value then we can expect it to run indefinitely at its rated value.

If we have a household circuit rate at 20 amps and a number of outlets rated at say 10A then it is possible to connect more load that the rated fuse value. If we connect say a 10A fan heater, a 5 amp one bar radiator (maybe in the next room), a 400 Watt plasma TV (about 2A), and some plug in mood lighting at say 1 A or less then all SHOULD be well. 10+5+2+1 = 18A. If somebody then turns on an electric jug rated at say 8A current rises to 26A. More than the 20A nominal value but less than the 2 x 20A = 40A we have said it will blow at. But if the plasma TV is off and is turned on suddenly the power supply input filters amy present a nearly pure capacitive load to the mains. The mains will be at random phase at TV turnon and usually a current spike will cause no problems. But on some random lucky (or unlucky) day the mains may be at the very peak of the mains cycle at turn on. The capacitor may have stored charge of opposite polarity from last turnoff leading to an even greater current spike. Add a possibly high mains voltage (as happens) and some heavy switching spikes from a nearby factory, or even domestic equipment (treadmill, welder, drill, sander, router, planer ...) Then load + capacitor spike + high mains + switching transient may lead to a very high short term load. And the fuse may decide enough is enough and melt. Or may not.

*Unlikely?*Is all the above likely to happen at once?
No. But as reported, the nuisance blowing happens only a few times a year. Ij the order of what is expected.

We could make the fuse rating higher (more amps)!
Yes. That is one solution. But the ability to react to moderate overloads is lost. Along with lack of protection may go loss of insurance, if the insurance loss assessors find a still intact 2 x 20A wire fuse in the smouldering ruins of your workshop.

Might want to see what the rating is, and check what you have connected. High current draw devices (like your microwave as you say above) can and will blow fuses/trip the breaker regularly. Try to put it on an alternate circuit if possible.

Every fuse has a number of important ratings:

1. The amount of current which a fuse is guaranteed to let flow indefinitely without any possibility of it blowing.
2. The amount of current which is guaranteed to cause the fuse to blow within a certain period of time.
3. The amount of current and voltage which the fuse is guaranteed to safely prevent from flowing once it blows (in some cases, the fuse's resistance once it blows will be sufficient to limit current to a safe limit provided its voltage rating is not exceeded; that is not always true, however, in the absence of external resistance).

If the amount of current that flows through a fuse is slightly more than the fuse can handle, the fusing material will slowly heat up to the point that it melts away, whereupon no more current will flow. If, however, the current that flows through it is much higher, some of the metal fusing material may vaporize. The metal vapor will itself be somewhat conductive, and the current flowing through that metal vapor may generate enough heat to vaporize more metal, increasing the current flow further. Given time, the metal vapor would dissipate and the circuit would open. The more current is driven through the fuse, however, the more heat will be generated before this occurs. If enough heat is generated, the pressure of the gas inside the fuse may increase sufficiently to cause failure of the containing material, sending pieces of it flying.

If the resistance of the circuit in which the fuse is placed exceeds the resistance of the vaporized metal, then an increase in the concentration of vaporized metal in the fuse will temporarily increase the amount of power dissipated in places other than the fuse (which might not be a good thing) but decrease the amount of power dissipated within the fuse itself. The fact that the fuse would be continuing to conduct might be a bad thing for the circuitry the fuse is supposed to be protecting, but reduction in power dissipation within the fuse would decrease the rate at which metal vaporized, reducing the likelihood of the fuse exploding. On the other hand, if most of the effective resistance in the circuit is within the fuse itself, the decrease in resistance caused by the metal vapor will increase the power dissipation within the fuse, thus causing more metal to be vaporized and reducing the resistance further.