Disconnection Time for IT system

[Cookie]

What is the max disconnection time for an IT system? I think it would be shorter in that an MCB would only be triggered after a fault on two individual circuits hence double the loop impedance?

 

[timhoward]

I think the answer its same as TN if there's a distributed neutral, or you double the OCPD trip current for the max Zs formula if there isn't.
Don't have BBB to hand so don't rely on that!
EDIT originally said half, sure that is wrong!

 

[Cookie]

Its alright Table 41.1 doesn't seem to mention IT? Or am I way off in my observation?

 

[timhoward]

Have book now!
See 411.6.5 note 1
And in fact while 411.6 section

 

[pc1966]

IT systems don't really have the same disconnection behaviour as the TT/TN set.

Basically if you have a single fault then nothing happens, except your insulation monitoring system starts to complain. This is perfect for an operating theatre or a ship at sea where the loss of power has serious implications.

Of course after the first fault it is no longer an 'IT' system but a TN system with some arbitrary choice of neutral based on which conductor had the fault. Then a 2nd earth fault on another of the live would trigger OCPD or RCD based disconnection.

Off hand I don't know if there is a spec for that, as normally on the first insulation fault you would be looking at scheduled repair work ASAP to sort it out and, perhaps, be on-guard for another fault. But based on it being TN-like after the 1st fault, and not a high imedance earth return like TT, I would expect it would be the 0.4s for UK based 230V nominal systems , etc.

 

[Julie.]

As highlighted above, in the case of a single live-earth fault there is no absolute disconnection time as the principle of the IT system is that normal operation may continue with a single fault. There are requirements for identifying such a fault condition.

In the case of a second fault, then this is where iec60364 and bs7671 differ.

There is a table in the iec which gives disconnection times for non-distributed neutral systems and distributed neutral systems based on voltage, for 230V this is 0.2s and 0.5s respectively.

In the bs (18th) part of 411.6.5 directs you to use the disconnection times as per TN - which is 0.4s.

 

[Cookie]

Julie, mind blown! I had not realized the difference, now I see it.

Can I ask- why does having the neutral distributed make a difference?What is the theory behind that?

 

[Cookie]

Also, would it be to much to ask for a pic of the table regarding distributed and none distributed neutrals?

 

[Julie.]

Also, would it be to much to ask for a pic of the table regarding distributed and none distributed neutrals?

A distributed neutral is for example 3ph +neutral or single phase + neutral

A non-distributed neutral is say 3 phase 3 wire, in this case as a double fault involves line conductors only it has the lower time - potentially larger current, and easier detection

 

[Cookie]

So basically TN vs IT? I can't picture at 133/230Y TN system needing the same disconection time as 230 volt IT. Both are none distributed neutral systems.

 

[pc1966]

So basically TN vs IT? I can't picture at 133/230Y TN system needing the same disconection time as 230 volt IT. Both are none distributed neutral systems.

The problem is after the 1st fault you might have 133V to earth (if N-E faulted) or you might have 230V to earth (if either of L1-L3 are faulted).

Realistically the higher phase-phase voltage should be your design criteria for shock protection disconnection times as it is more likely to happen under fault condictions than a phase-neutral voltage (that a TN system's parameters are designed around).

 

[Cookie]

Right, but picture a 133/230Y TN system. Highest voltage to earth will be 133 volts no matter the network conditions.

 

[pc1966]

Right, but picture a 133/230Y TN system. Highest voltage to earth will be 133 volts no matter the network conditions.

For the TN, yes, but for a distributed neutral IT system based on the same star/wye voltages you could end up with the 230V to earth (probably in half the double-fault cases but not checked).

 

[pc1966]

Yes, for a double-fault on a 230/400V 3P+N IT system half the non-zero cases are 230V to earth and the other half 400V, as seen here:

1st fault/2nd fault​	N​	L1​	L2​	L3​
N​	0​	230​	230​	230​
L1​	230​	0​	400​	400​
L3​	230​	400​	0​	400​
L3​	230​	400​	400​	0​

 

[Cookie]

Makes sense now, thank you!