[vhfpmr (OP)]

Has anyone ever seen any stats on what proportion of electrical failures blow the fuse/breaker supplying them vs those that don't? (It's the incidence proportions I'm looking for, not a list of fault scenarios.)

Googling this, I can't get past the endless Noddy's guides to how a fuse works, so I tried Chat GPT for what it's worth. After listing potential faults, it claimed that 40-70% of failures will trip the protection, but when I asked for a source for that it said "it's challenging to provide precise proportions of failures that do or do not cause protective devices to trip", and claimed it's just "general knowledge".

I then asked about safety critical failures as opposed to failures in general, and it claimed the trip rate rose to 60-90%, but with exactly the same disclaimer.

(Digressing a bit, it did mention an IEEE study that found 50% of some brands of circuit breaker don't trip at their rated current, even when new.)

[Petrochemicals]

Depends what you mean by failure, a wire fuse has a set tolerance but this is obviously open to different preciseness, like any other  electrical component. Wire fuse  boards have specifications for the reason you describe.

I would hope that the wire in a fuse was 100 percent effective at preventing catastrophic overload.

[Bored chemist]

Most of the fuses I see are in the plug, and they are designed to protect the wiring.
As far as I'm aware, they usually manage that. It's hard to see how they would fail to do so.

[alancalverd]

Usual reason is that someone has replaced a 3A fuse with a 13 "because it kept blowing". The beauty of the clumsy British mains plug is its irreversibility, so if you get a live-to-earth fault in the machine, the fuse blows and renders the machine safe. Naughty foreign plugs can be reversed so they put two fuses in medical equipment and a fault can leave the casing of the machine live.

Anyway, apropos the OP, fuse blowing is much less frequent than residual current faults. In addition to providing overcurrent protection most rccbs trip out at around 5 mA earth current. This at least alerts you to a fault in the wiring and also prevents the (earthed) metal bits of the machine maintaining more than about 5V above earth potential, so makes the system virtually shockproof.

Mcbs have different overcurrent trip speeds, from milliseconds to several seconds (motor starting currents can be many times the running current).

In my experience most trips  are either due to  an intermittent residual current fault in the supply cable (rats eating an old cable, or a new build where someone has driven a nail through the cable) or an inadequately rated (B for cookers and most domestic ring mains) mcb installed where a C or D was required for an x-ray machine.   Actual machine faults? Probably 5  or 10 in a lifetime of installing and testing all sorts of things, two of which were domestic cookers.

[paul cotter]

5ma? All domestic rcds I have come across are rated at 30ma residual. 30ma will delver a painful shock.

[evan_au]

I can't give percentages, but most failures in electronic appliances should not blow the fuse in the fuse box; the fault should be contained within the appliance and not affect any other devices on the same mains circuit.

Because mains electricity is potentially lethal to people, and could set buildings alight, the electrical industry has adopted a "defense in depth" approach. The fuse/Circuit Breaker/Residual Current Device in the fuse box is not the only layer of protection.

Some of these protective layers include:
- Double-Insulation: At least 2 layers of insulation between users and potentially lethal voltages. Mostly used for small domestic and medical appliances; not so much used for large domestic and industrial devices.
- Earthing: A metal frame is connected to an Earth stake or Neutral point, so it limits voltages in case of fault, and intentionally blows the fuse or trips the circuit breaker in case of an electrical fault. Used for metal appliances.
- Safety testing: This includes subjecting appliances to foreseeable failure modes (at least on paper), and showing that the device shuts down safely (perhaps with a fuse inside the appliance).
- Fuse in the Plug: This seems to be primarily used for domestic appliances in the UK; I have not seen it used in other countries.
- Per-Circuit Fuse/Breaker in the fuse box: The house is divided into separate electrical circuits, eg for lights, and for power points in various parts of the house; large appliances might have their own circuit (eg stove, air-conditioning). Refrigerators seem to be on a separate circuit, as their moist environment is prone to tripping RCDs. This means that a failure of one appliance will not take out other circuits.
- Residual Current Devices (RCD): These detect small electrical currents to Earth; this can happen if a person (or a rat) touches a live circuit. They are designed to cut off power before the current could kill a human (I don't know how a hungry rat would fare...). Presumably, this could also be tripped by a fire?
- Mains Fuse: In addition, there is a fuse between the house and the street wiring; this should never be needed, due to the previous layers of protection. However, in severe cases (eg a house fire), this is intended to protect the power supply to other houses in the street.
- Circuit Breakers and Fuses in the Distribution Network: Your electrical supplier will install protection devices throughout their network, for example at the distribution transformer supplying your street, and the much larger transformer feeding your suburb. So if a truck knocks down a power pole, the transformer is protected.

[alancalverd]

5ma? All domestic rcds I have come across are rated at 30ma residual. 30ma will delver a painful shock.

I think you are correct! I've been looking at limits for medical systems.

[paul cotter]

Indeed, one would not want a urethrascope delivering 25ma and failing to institute a trip. Line filters for medical equipment are required to have minimal leakage current, I can't remember but I think it is less than 1ma. I have seen some industrial rcds rated at 300ma- I am not sure on what criterion they are designed but it cannot be human safety at that current.

[Petrochemicals]

I have had shocks from cables that do not even trip the RCD, it must mean I am electrically balanced. They do only trip if the forward and return currents are out of line, earth leakage being what the stock answer, say you are holding a water pipe in the right hand and touch a live with the left, ecectricity across the heart stopping it, which is the danger. I have had rcd trip at a bulb going or similar, this doesn't necessarily mean leakage and faulty device, just that at a certain point the current was out of parity. I suppose the 5mm in-between the live and neutral conductors across my skin doesn't create enough of a ripple in the current, not sure what the full arms pan would do.

The modern equivalent of a physical fuse is the circuit breaker, but they are whole amp rated in the house.

[paul cotter]

If you feel a shock then you are passing some of the current from the live or hot conductor which would normally return through the neutral. If the rcd fails to trip either it is faulty or your shock is below the 30ma trip level. Being "electrically balanced" is meaningless.

[Petrochemicals]

If you feel a shock then you are passing some of the current from the live or hot conductor which would normally return through the neutral. If the rcd fails to trip either it is faulty or your shock is below the 30ma trip level. Being "electrically balanced" is meaningless.

Not really meaningless,  assuming touch the return too, but that is a good point if I am not touching the return does it actually mean a fault at 30ma ?

[vhfpmr (OP)]

As above, it was data I was looking for, I'm familiar with the scenarios that do and don't blow fuses, not least from explaining them to others. Fuses fully protect supply side circuits, but can only ever partially protect load side circuits. By "fuse/breakers" I was referring to fuses or the MCBs that replaced them, not RCDs, and by implication safety was really referring to fire, since that's mainly what a fuse can protect against. There aren't many people able to withstand enough current to blow a fuse.

In terms of value for money, fuses are pretty good, even if the proportion of faults they cover is limited.

foreign plugs can be reversed so they put two fuses in medical equipment and a fault can leave the casing of the machine live

TN-C-S supplies also leave appliance casing live in the event of a break in the neutral conductor on the service cable, but all exposed metalwork in the building is supposed to be connected with equipotential bonding so that a potential victim can't find two different potentials to touch simultaneously in that scenario.

If you feel a shock then you are passing some of the current from the live or hot conductor which would normally return through the neutral. If the rcd fails to trip either it is faulty or your shock is below the 30ma trip level.

I was a bit surprised to find that I could feel the leakage from my double insulated HiFi when I bought it, so I measured it, and found it was well within spec. Forty years on, my fingers aren't sensitive enough to feel it any more. I had an electric blanket that was leaking enough to feel once.

[Petrochemicals]

There aren't many people able to withstand enough current to blow a fuse.

The lighting circuits where capped out at 2A years ago and where not considered a lethal threat, hence why people played fast and loose with them. I suppose they didn't cater for pace makers.

but all exposed metalwork in the building is supposed to be connected with equipotential bonding

depends on the system design.

[paul cotter]

2amps is most definitely lethal, pacemaker or not.

[alancalverd]

  The leakage current limits for medical equipment are specified by IEC 60601-1 and include the following:

Earth Leakage Current: Typically limited to 0.5 mA for equipment with direct patient contact.

Patient Leakage Current: Limited to 10 ?A for applied parts that are in direct contact with the patient.

Chassis Leakage Current: A practical limit is often set at 500 ?A.

Type B Equipment: Should not exceed 0.1 mA for patient leakage current.

These limits are crucial for ensuring the safety and effectiveness of medical device

The human body is conventionally represented for test purposes as 1000 ohm in parallel with 1 microfarad. The IEC limits are, IIRC, set at one tenth of whatever will cause severe temporary but nonlethal effects.

[alancalverd]

all exposed metalwork in the building is supposed to be connected with equipotential bonding

We had an interesting problem with an early digital x-ray system in a hospital tower block. It was impossible to get a clear image between 11 am and 1 pm. Problem turned out to be that the 3-phase kitchen equipment was bonded to the steel frame at a different point from the ring main earth. Lunchtime produced anything up to 10 V differentials......

[paul cotter]

Connecting two "earths" with a potential difference can cause problems with circulating currents.

[Petrochemicals]

Digressing a bit, it did mention an IEEE study that found 50% of some brands of circuit breaker don't trip at their rated current, even when new.)

I seem to remember that a ring main circuit breaker is 32 amps, which is less that 3 3kw appliances, a kettle, a washing machine dishwasher and a tumble dryer running, someone flips the kettle or puts the oven on and you are at 40A+. I would think it is quite a regular occurance that the rated current is exceeded, especially morning or night.

[Petrochemicals]

2amps is most definitely lethal, pacemaker or not.

All dependant again on the path, stands a chance most people would only loose a finger. Remember people do survive lightning strikes.

[paul cotter]

Yes of course the path available to make a circuit is the most important factor. Wearing trainers and standing on a wooden pallet I once accidently brushed against the anode(plate) cap of a valve(vacuum tube) at ~ 2kv with my hand. I felt absolutely nothing but I heard a faint crackling sound. The psu in question was capable of a good 5amps. As regard lightning strikes some people will survive a direct hit but most will not and those that survive are rarely the same again with multiple problems. A lot of cases involve a direct hit in the vicinity which can still induce catastrophic currents in an individual but way less than a direct hit.