IT system or TN system? I believe Norway mostly uses IT in distribution. Is one advantage of TN lower cost and less equipment? I think both earthing arrangements have advantages and disadvantages for different applications.

What you describe is an IT system (insulated from Terra-firma) It certainly can be done, and in some cases it has advantages, and indeed is used for UK shaver sockets in bathrooms, and some types of specialist medical equipment.

However the shaver socket only supplies one shaver, and the medical kit is very carefully maintained and monitored when in use

So why do we not do it more commonly ? Well the answer is all to do with automatic disconnection of supply (ADS).

If you do cannot detect  a fault perhaps from one phase to the case of  an appliance, . then nothing is disconnected, and the presence of a second fault between another phase and earth makes things suddenly lethal with no warning.

When your substation feeds many houses each with some tens to up to perhaps a  hundred or so items connected for years, the chances of at least one  fault being present at some time become quite high.

With the N-E connection, you can detect and automatically isolate the faulty equipment with fuses if the fault loop is low impedance, and with an RCD if it is not. Neither form of ADS will detect faults on an IT system.

Note also that there will always be a rather indeterminate path between every conductor in the system and earth, as there is capacitance and insulation resistance to be considered, so in effect things float about, unless they are tied down and there is a voltage but not a well defined one. And when the installation is large then the current available via these unavoidable leakage paths becomes enough to make touching the conductors  dangerous anyway.

However it is only convention that we earth the central star point of a 3 phase transformer - you could earth one corner, or as sometimes seen in the states, the centre of one side. The 3 phases still have the same voltages between them, and motors etc work just fine but the voltages and phase angles relative  to earth are all over the place..

Mike.

I suppose that another way of putting this is why not put a whole (SP) installation on an isolating transformer and do away with CPCs?

So let's imagine that the line in your toaster comes adrift and touches the casing, and the neutral in your kettle comes adrift and ditto. You pop in your slice of bread whilst turning on the kettle and you are toast.

But in that case you'd be creating a short circuit from phase to neutral which would blow/ trip the fuse or protective device would it not?

no nothing would blow - there would just be 230V between the two metal objects, and if you connect yourself between them, your body resistance of a few k ohms  looks like a small extra legitimate load, nothing trips. There is no CPC connecting the cases  remember.  The only time anything trips is when there is a near short between the cases of the two defective devices.

Mike

Yes, good point, my bad. 

If I poke my finger into the live terminal of a 230v socket, the line current travels through me, through the floor (earth), back through the earth to the supply transformer (because the transformer star point is also tied down to the general mass of earth) the fault currents travel back into the transformer star point and hopefully the completed path is of low enough impedance to be able to blow or trip the protective device serving the socket I touched.

Almost, but not quite. The human body itself has considerable resistance (varies between individuals, but typically in the region of 1000 Ohms) - so the 'path of low enough resistance' only works for faults between line conductors and deliberately earthed metalwork (exposed-conductive-parts) - people are then given some protection if touching that metalwork while in contact with the general mass of the earth (or other earthed metalwork). What used to be called 'protection against indirect contact'. Putting your finger on a line terminal (or other hazardous live part) is much more difficult - traditionally the risks were avoided by making it difficult to touch - e.g. by enclosures of insulating material or earthed metal - nowadays we can provide some additional protection as it's called by 30mA RCDs - which spot the small currents leaving the normal circuit.

After all, this is exactly why we use SELV transformers in bathrooms no?

Indeed. Usually comes under the heading of "separated" rather than IT if there's nothing earthed on the load/consumer side. As mentioned above it's actually quite hard to achieve sufficient separation - cable insulation is never going to be totally perfect (especially buried ones) and will have considerable capacitance to Earth (as we use a.c.) The regs recognise separated systems, but put a limit of 500m and 100,000Vm (so around 400m for 240V) on the wiring system - which probably counts out even a normal dwelling.

   - Andy.

That is before you have the several hundred metres of thinly insulated wire in the average motor or transformer wrapped around metal bits that may or may not be connected to the chassis, where the winding to core capacitance in hand held things are of order of a few nanofards unless extra thick insulation is used between windings and frame.

Ordinary twin and earth manages 50-100picoFarads per metre between adjacent cores in the common sizes, and more like 10-20pF per meter between the cores and metal tray or trunking, less to things like basket that are mostly holes and form a minimal second plate to the capacitor more when pressed to solid metal.

Mike.

Adam Seddon said:

Is that correct or not?

So following on from that, my question is: why cant we just isolate all transformers from earth and remove the risk of a "line to earth" shock? After all, this is exactly why we use SELV transformers in bathrooms no???? 

Obviously you could still get a line to neutral shock, but that's got nothing to do with the question I'm asking here. 

Yes, you are correct, that is how electrical separation and IT system works.

The problem with an unearthed system, is that on the first fault of the system to something conductive connected to Earth goes undetected ... making it a TN system effectively, and it's thus the second fault that kills.

Hence the additional requirements in BS 7671 for IT systems ... and limitation of 1 piece of equipment supplied by electrical separation (unless the installation is under constant supervision of skilled and/or instructed persons).

I still wonder about the actual science behind perceived electric shocks,at college(in Stevenage) we was told its the resistance between you and the mass of earth , but in my thick rubber sole boots on a carpet on a wooden floor this can’t be true.

i have measured a few pico farads between me and earth (though this was measured at 1000 hertz) so I assume this is the current path?