Understanding Electric Shock in Floating-Neutral and Floating-Ground Scenarios

[zabala9]

Hello, let's say the AC generator and the washing machine, with a live metal frame due to a fault, are floating-neutral and floating-ground. These are many combinations where both get and get not bonded/grounded. Can anybody tell us in which case the person gets an electric shock when touches the metal frame, and why? The person is standing on the soil with bare feet, and with shoes (to put both cases)
This isn't a homework, it is just that there are a lot of different opinions on the internet and some confusion in me and many people. I added with/without RCD/GFCI breaker cases to make this more complete
thanks

 

[Baluncore]

It will depend on how the person is also grounded.
None of your diagrams show a ground for the person, so there would be no ground current or shock.

 

[zabala9]

thanks, I added it

 

[berkeman]

AC generator and the washing machine, with a live metal frame due to a fault

Due to a double-fault, not just a single-fault. Can you comment why it would take a double-fault (at least in the US) to create a shock hazard like this?

 

[zabala9]

double-fault? can you explain us what is that? only the washing machine has its frame energized in all the cases shown, as can be seen in the diagrams

 

[berkeman]

double-fault? can you explain us what is that?

Safety regulations for appliance design and installation have a goal of not creating a dangerous situation as a result of only a single fault. In the US, it is Underwriters Laboratories (UL) that publishes various design standards, and does testing and qualification of devices to those safety standards.

So for AC Mains powered appliances and instruments, there are regulations for how you handle the AC Mains wiring and the grounding of the devices. For metal body appliances like washing machines, a 3-prong power cord is used so that Earth Ground is available inside the machine. That Earth ground wire must go directly to the metal body of the machine, and be connected with a solid, reliable connection (like a metal stud, washer and locknut). That ensures that all externally-accessible metal on the machine is reliably Earth grounded.

In addition, the AC Mains Hot wire that comes into the machine must go immediately through a fuse and a switch of some kind before going into the power supply. Those connections also need to be very reliable (often made with spade lug type connectors).

All of this is done so that a single fault (a single thing going wrong) cannot energize the metal enclosure of the machine and create a shock hazard. If the Hot wire breaks or falls off of a connection for some reason, if that wire does contact the metal enclosure, it creates a short through the fuse and the fuse blows quickly enough so that there is no shock hazard. If the Earth Ground wire connection breaks or falls off for some reason, the Hot wire is still reliably bonded in the system, so there is no path to energize the floating metal chassis.

So you can see how with this system of safety rules, it takes a double-fault to energize the metal chassis. That is why it is so unlikely and hardly ever happens.

 

[zabala9]

ok berkeman thank you for explain us that the washing machine was a bad example. A toaster 'live' frame would be ok in order to go back to the main point of the question.. in which case the person is getting an electric shock?

​

 

[berkeman]

A toaster 'live' frame would be ok in order to go back to the main point of the question..

Toasters are usually double-insulated Class II devices with 2-prong AC Mains plugs...

https://en.wikipedia.org/wiki/Appliance_classes

What exactly are you trying to ask?

 

[zabala9]

Toasters are usually double-insulated Class II devices with 2-prong AC Mains plugs...

so it can't have a faulty line wire inside that could energize the frame?? can you tell us a device which metal frame can be energized by a faulty wire?

 

[berkeman]

so it can't have a faulty line wire inside that could energize the frame?? can you tell us a device which metal frame can be energized by a faulty wire?

Both the Class II doubly-insulated devices and the ones I described earlier require double-faults to energize external metal on products. What kind of faults are you concerned with?

 

[zabala9]

ok, sorry for trying to learn

 

[berkeman]

ok, sorry for trying to learn

There's nothing wrong with that.

It's most important to understand when you need to use a GFCI/RCD in the AC Mains connection. Do you understand why it's needed in outdoor applications and in indoor applications near sinks and bathtubs, etc.? After all, throwing a powered toaster into a full bathtub is a single fault...

 

[DaveE]

In addition, the AC Mains Hot wire that comes into the machine must go immediately through a fuse and a switch of some kind before going into the power supply. Those connections also need to be very reliable (often made with spade lug type connectors).

All of this is done so that a single fault (a single thing going wrong) cannot energize the metal enclosure of the machine and create a shock hazard.

In common safety standards (IEC-950 et. al.), fuses (and circuit breakers, etc.) are used for fire protection, not for protection against hazardous voltages. There are single fault scenarios where fuses simply aren't reliable or fast enough to deenergize the fault. This is unlike GFCI (or RCD*) devices which are designed specifically to be fast and sensitive enough, and to measure the relevant parameter (fault current) to save humans.

Thus the ground wire(s) are required to have good mechanical construction and sufficient conductivity to reduce the accessible voltage to safe levels in a single fault condition essentially forever. In this scenario, it isn't uncommon for things to get hot and maybe start a fire. This is where fuses come into play; as well as other requirements like accessible temperature, insulation ratings etc.

*Yep, I'm American! RCD confuses me, but I think I've fixed that by pretending y'all really meant to say GFCI, LOL.

PS: The safety standards I'm familiar with (certainly not all of them!) will require safe accessible voltages, currents, temperatures, etc. under single fault conditions. These are typically written as performance standards, not construction standards.

This means that you can use fuses, or whatever else you like, to prevent a specific hazard. But, you'll have to convince an inspector, usually with LOTS of testing, to prove your design is safe. Of course everyone, designers and inspectors, resort to standard solutions to avoid the complexity. This I all learned the hard way back in the day getting safety approvals for huge Ion Lasers, which are really difficult from a "normal construction methods" standpoint. The practical cost was we had to do lots of fault testing and find really knowledgeable (i.e. rare and expensive**) inspectors to get our stuff done. Your normal "new grad safety guy" can't handle it, they don't really know the real requirements. All they every see are standard products.

None the less, if you've ever waded through the trip characteristics of fuses or circuit breakers with a "worst case" mind set, you'll soon give up on them for doing anything quickly.

** Short answer, back in the day, they probably work for TUV, definitely not ever, ever, ever UL.

 

[berkeman]

In common safety standards (IEC-950 et. al.), fuses (and circuit breakers, etc.) are used for fire protection, not for protection against hazardous voltages. There are single fault scenarios where fuses simply aren't reliable or fast enough to deenergize the fault. This is unlike GFCI (or RCD*) devices which are designed specifically to be fast and sensitive enough, and to measure the relevant parameter (fault current) to save humans.

Thus the ground wire(s) are required to have good mechanical construction and sufficient conductivity to reduce the accessible voltage to safe levels in a single fault condition essentially forever. In this scenario, it isn't uncommon for things to get hot and maybe start a fire.

You are probably right about that; thanks for the clarification. In my experience, I've had fuses go off like a gunshot (so very fast), but I can see how a skipping wire touching the inside of the metal enclosure might not generate enough heat in the fuse to cause it to blow (especially with a slow-blow fuse type). And I also agree that even given that, the strong ground wire bonding to the chassis should help to keep the voltage on the sheet metal below SELV so not a shock hazard during the fault.

 

[DaveE]

You are probably right about that; thanks for the clarification. In my experience, I've had fuses go off like a gunshot (so very fast), but I can see how a skipping wire touching the inside of the metal enclosure might not generate enough heat in the fuse to cause it to blow (especially with a slow-blow fuse type). And I also agree that even given that, the strong ground wire bonding to the chassis should help to keep the voltage on the sheet metal below SELV so not a shock hazard during the fault.

The bonding system (wire gauge, mostly) then, basically, has two requirements for conduction:
1) It must keep accessible voltages within safe limits. This requirement is typically easy to meet.
2) It must survive worst case current flow. This is the connection to fuse size. The fuse must keep the ground wires from failing (in addition to fire protection).

 

[Tom.G]

If I recall correctly, the National Electrical Code here in the USA requires the Safety Ground wire (Green in the USA, Green&Yellow in Europe) to be Not Smaller Than The Current Supply Conductors.

 

[DaveE]

If I recall correctly, the National Electrical Code her in the USA requires the Safety Ground wire (Green in the USA, Green&Yellow in Europe) to be Not Smaller Than The Current Supply Conductors.

The NEC is a reference for local building code laws. It is very much a "safety by required construction" sort of standard as oppose to the product safety standards which are mostly performance requirements.

This makes sense because builders don't do mass production and can't be expected to test and certify each design, also, buildings are usually modified by someone else during their life, and those people need to know what to expect and how to conform to the safety plan.

OTOH, most manufactured products have exactly the opposite situation. So detailing construction requirements is too restrictive, or just doesn't make sense.

 

[Guineafowl]

Here is an example of a single fault enlivening any earthed metal appliance.

Imagine the blue CNE/PEN conductor breaks to the left of the link to the earthing conductor. The equipment case is then connected to live through the load and N-E link.

An RCD/GFCI won’t operate in this situation.

This is a very common earthing system in the UK, and is also known as TN-C-S. The risk is commonly mitigated between transformer and installation by using split concentric cable, which has the CNE/PEN conductor distributed as multiple wires around the other conductors, reducing the likelihood of a complete break.

It’s also a concern with EV chargers, since a broken PEN will make the car‘s bodywork live.

 

[Tom.G]

Here in the USA, a single phase GFI (Ground Fault Interruptor) senses a current unbalance[\i] in the power conductors. If the currents are not the same, it disconnects the power. I have no idea how the multi-phase GFIs operate.

 

[DaveE]

I have no idea how the multi-phase GFIs operate.

The same way as single phase. You add up all of the "going away" currents and subtract the "coming back" currents. If it isn't zero, someone might be getting killed. This is (or was) usually done with separate windings on a transformer, similar to a common mode filter.