Discrimination and correct implementation of Zs

[1ManRiot]

Scenario:
3 phase DNO with 100A fuses into a control unit feeding 2x 5kVA txs with combined load of 40A.

We're adding 2 more 5kVA txs, one with 14A, the other 21A.


What I've come up with so far:
63A fuse consumer side will need replacing with an 80A BS 88-2 (gG) fuse (max Z[SUB]s[/SUB] 0.57Ω).

Existing 70mm[SUP]2[/SUP] 3-core using reference method D has a current carrying capacity of 173A with an existing load of 40.3A and an estimated 55A in-rush current.

Points 20A
Load 14A
Protection 20A C-Type BS 60898 providing a required Z[SUB]s[/SUB] of 1.15Ω
Cable 144m of 16mm[SUP]2[/SUP] 3-core at 2.30mΩ/m = 144 x 2.3x10[SUP]-3[/SUP] = 0.3312Ω which satisfies the protection Z[SUB]s[/SUB]
Volt-Drop VD = 2.4mV x 14A x 144m = 4.8384V
230V - 4.8384V = 225.1616V (230V – 5% = 218.5V)

Points 20B
Load 21.2A
Protection 25A C-Type BS 60898 providing a required Z[SUB]s[/SUB] of 0.92Ω
Cable 100m of 16mm[SUP]2[/SUP] 3-core at 2.30mΩ/m = 100 x 2.3x10[SUP]-3[/SUP] = 0.23Ω which satisfies the protection Z[SUB]s[/SUB]
Volt-Drop VD = 2.4mV x 21.2A x 100m = 5.088V
230V – 5.088V = 224.912V (230V – 5% = 218.5V)

Does this look ok so far? I know 16mm2 is massive overkill but that's what has been specified by the powers that be.

I'm a little confused at the Zs aspect of it though, in how does the Zs of each circuit tie in with the entire system Zs? Is it a case that so long as the Ib <= In <= Iz < = It) is correctly calculated for each stage, with sufficient discrimination applied at each level, Zs will be correct throughout the system?

If that doesn't make much sense it's because I'm still new to this and could have worded it incorrectly!

(Before the "don't know what you're doing" brigade chip in, someone far more experienced will be going over this in its entirety, I'm merely trying to learn for myself)

 

[Guest55]

i'll "attempt" to untangle your understandings..........

firstly , there is no such thing as a system Zs , you need to evaluate each circuits Zs seperately.
secondly , following Ib > In > It does not guarantee discrimination between devices.
for example ,
if you have a 20a mcb in a CU protecting a socket radial , the upstream device feeding the CU needs to be double the rating of the 20a mcb to ensure discrimination during fault conditions , so you end up with a 40a mcb feeding the CU.

if the max Zs of the 20a mcb is 1.5 ohms , and the R1+R2 of the socket radial is 0.5 ohms , this means that the Zs at the CU cannot be higher than 1 ohm.
once you know what the max Zs can be at the CU , you can then work out what size cable you need to supply the CU.

in essence you start at the furthest point of your final circuit , then start working back through the DB's and submains , back towards the supply origin.
it sounds a bit counter-intuitive to start at the end and work backwards , but its the only way to get the system to "fit together" correctly.

you can only go so far trying to explain all this on the internet but i hope it gives you a bit of a clue.

 

[1ManRiot]

firstly , there is no such thing as a system Zs , you need to evaluate each circuits Zs seperately.

Ok I think that's cool, I think that's what I've been doing! Basically the DNO protection will see the control cubicle's (or CU's) total load, it doesn't matter what the breakdown of the circuits after that, so long as the total meets the protection requirements at the DNO (ie, less than 80A)?

in essence you start at the furthest point of your final circuit , then start working back through the DB's and submains , back towards the supply origin.
it sounds a bit counter-intuitive to start at the end and work backwards , but its the only way to get the system to "fit together" correctly.

Yeah that makes sense, that's what I was trying to do I think, I wanted to ensure that I wasn't asking too much of the final circuits with regard to the CU and then to the DNO and dragging the Zs up too high as to cause an issue, but it seems that's not how it works, as it is staged/levelled...