The maximum permissible disconnection time is 0.4 s in TN system. Why and from where this value (0.4 s) is obtained?

There's a famous graph )from IEC 60479) of how long a (typical) human body can withstand various currents for (for various different results, from nothing (AC-1) to a slight tingle (AC-2) right up to various probabilities of death (AC-4)):




To make that useful for protecting a human from electric shock you need to make a few assumptions - e.g. the resistance of the body between contact points (e.g. hand and foot) but if you assume 1kΩ you can then read milliamps on the graph for volts across the body. In a TN system with a fault of neglible impedance between L and PE you'd get a voltage of about half the nominal voltage at the point of the fault (half dropped along the line conductor, half along the PE) - looking up about 115V (115mA on the graph) and staying out of the AC-4 regions should give you a time of about 400ms (or 0.4s if you prefer).


For TT systems the PE part of the loop (via the customer's electrode) has a much higher resistance than the line conductor, so we can't assume a touch voltage as low as half line voltage - indeed in worst cases it could be almost the full line voltage - 230V say - so looking up 230mA should give you a result closer to 0.2s.


(I just pulled that above image randonly from the internet - there do seem to be a few different versions about that differ slightly (but sometimes significantly) in the details - so don't get too worried if things don't seems to line up precisely.)


  - Andy.

Andy


Where did you get that colour graph if it is not a rude question as the one in 60479 is B&W.


You have mail by the way.

And the reason it is that shape, and all the action is in the region of a few hundred milliseconds,  is to do with the human heartbeat timing, and the chances of a current passing through the body putting the heart muscles into an irreversible fibrillation, so that it no longer beats properly.

A current surge that is very short compared to a heart beat is can be much higher before it has the same effect as a lower current of long duration.

In slang, you need to get the fault current interrupted in half a heartbeat to give the victim a good chance of survival.

Hidden behind this are various assumptions about hand to foot shocks, and how good the contact really is, most of the time a glancing contact is a lot higher resistance than a kilohm, so a lot of shocks that look like they ought to kill just cause some bad language and a bit of a flinch,  these figures assume good contact.

The limit of  50V for ELV has another assumption that the current is low enough that it needs no ADS,  but beware - being cut on live metal so you lose the skin protection, or even just being wet with seawater can lower your resistance, and means this is not true,  and a lower 'safe' voltage, like 25, for example, may have been a better choice.

Once upon a time some confined working regs acknowledged this, and recommended a lower voltage for tools for crawling into wet pipes etc.


As a final note, the volts and currents in a defibrillator are so large, the idea is to clench the muscles up, so the heart is actually stopped in a known state, and the idea is that when the current stops the heart re-starts in correct rhythm,  a sort or reset. Works often enough to be worth trying anyway.

What about the 5s disconnection time where did that come from? I know GB can answer that one but may not want to!

Where did you get that colour graph if it is not a rude question as the one in 60479 is B&W

Google found it here for me: https://commons.wikimedia.org/wiki/File:IEC_TS_60479-1_electric_shock_graph.svg

 

What about the 5s disconnection time where did that come from?

The legend I heard was that it dated from (smoke filled rooms) when the regs committee first started discussing disconnection times and one member did a quick back of an envelope calculation based on what he knew about the circuits where he worked in his normal day job - and reckoned that they'd have to permit up to 5s if his system was going to continue to comply.


There's probably a co-incidenal relationship between most fuse characteristics and an acceptable voltage drop - which usually means that if the voltage drop is acceptable the loop impedance will be such that disconnection time will probably happen within about 5s too.


   - Andy.

Andy


Yes that is the aprocaphable tale I heard about the 5s disconnection time. 


I do do not know if there is any written requirement to lay rubber matting in front of LV panels but I think it may be based on grasping a nice shiny chrome operating handle and closing the switch on to a fault condition and waiting for 5s before ADS.

5 sconds is a long time to get even a half mains kick, so for that matter is 0.4 seconds. Even RCDs let you get a real kick too!

For 5 seconds, it is as good as 'forever' in terms of being  many cycles of heartbeat, you need the current to be below 30mA - which should be seen as the upper limit for 'probably survivable' certainly not enjoyable.

At 30mA, actualy rather below, muscles spasm, and you can expect bruising internal and external from the muscle movement. The body can take more at higher frequencies, but it needs to be carefully controlled. Equally, with a current path that does not include the torso, you can survive considerably greater currents, so long as heart and head are avoided.


The two viable  ways to reduce the current to a value ssafe for linger exposure are either to increase the resistance (Johns matting, or in my lab world, a waxed wooden floor or linoleum, or when out in the field rubber soled boots..), or to ensure that both ends of the human being are at more or less the same voltage, live or not  (the point of equipotential bonding.)


The RCD does not limit the current - the supply voltage and and the body conspire do that. The RCD can only limit the time of exposure, hopefully to a duration short enough that  even if the current is  over 100mA you have a reasonably good chance of surviving. Without an RCD, you can add burning, cooking and higher risk of death to the symptoms as above.

Testing a 30mA RCD at 5 times In, is not enough to be truly sure of this.

"Testing a 30mA RCD at 5 times In, is not enough to be truly sure of this " agreed