Generator Protection and PSCC values

[EDoubleD]

I was asked to go and take a look at a generator set up by a marquee company yesterday to do some testing.

It was a brand new JCB 65QS with 209hrs on it (58.7kva prime). The generator was fitted with a socket pack from the bus bar with 3 x 16a on 30ma RCBO, 1 x 32/3 on a C curve MCB and a 63/3 C curve. It has an ELR set to 300ma, 200ms.

It was set up with a 16mm 30m lead from the 63/3 socket pack to a distro board with RCBO's on 16/32 circuits....so far so good.

On beginning testing I was getting 0.96ohms at the ceeform socket for Ze (L-E) and 0.71ohms L-N. As the supply was on a C curve 63a mcb I won't make my ADS in 5secs with a PSCC below the required 630amps ( max 0.35 table 5.9). Further down the system everything else was on RCD's and mostly 16a C curve. The overall system had fairly short cable runs so I was making all the required times/PSCC at the 3 phase distro board.

The issue in my mind is that in the event of a short I'm not making the required fire protection on the 63/3 feeder. On speaking to someone far more knowledgable than me (and doing some research online) I can see that the generator isn't able to deliver the fault current anyway and that the method of protection would be the generator stalling, is this correct? It seems kind of crazy to me?

I've looked into methods of calculating the max PSCC the alternator can deliver so I can see what the actiual figure is:

This requires a figure for generator subtransient reactance but I don't understand where to get this figure from. I probably don't have a full handle on the detail here so hoping someone more knowledgable can explain this to me and what the general rules are for signing off small generators under 7909 where you're not making the required regs via testing at Ze for your feeded ADS. I've got a copy of the James Eade temporary power systems and it doesn't go into detail on this specfic scenario.

I suspect one of the issues I have is a lack of understanding about the subject. I feel like some more training is needed....they don't teach you half this stuff on a basic 7909 course!!! If anyone who knows about this indepth can hekp to explain it would be really appreciated.

Cheers

 

[pc1966]

It was a brand new JCB 65QS with 209hrs on it (58.7kva prime). The generator was fitted with a socket pack from the bus bar with 3 x 16a on 30ma RCBO, 1 x 32/3 on a C curve MCB and a 63/3 C curve. It has an ELR set to 300ma, 200ms.

You would meet ADS on with the RCBO or the ELR anyway.

The fault current / adiabatic implications are not likely to be anything like the Zs measurement implies as I suspect the generator would simply stall if asked to deliver 200+ amps!

 

[cireland]

I was asked to go and take a look at a generator set up by a marquee company yesterday to do some testing.

It was a brand new JCB 65QS with 209hrs on it (58.7kva prime). The generator was fitted with a socket pack from the bus bar with 3 x 16a on 30ma RCBO, 1 x 32/3 on a C curve MCB and a 63/3 C curve. It has an ELR set to 300ma, 200ms.

It was set up with a 16mm 30m lead from the 63/3 socket pack to a distro board with RCBO's on 16/32 circuits....so far so good.

On beginning testing I was getting 0.96ohms at the ceeform socket for Ze (L-E) and 0.71ohms L-N. As the supply was on a C curve 63a mcb I won't make my ADS in 5secs with a PSCC below the required 630amps ( max 0.35 table 5.9). Further down the system everything else was on RCD's and mostly 16a C curve. The overall system had fairly short cable runs so I was making all the required times/PSCC at the 3 phase distro board.

The issue in my mind is that in the event of a short I'm not making the required fire protection on the 63/3 feeder. On speaking to someone far more knowledgable than me (and doing some research online) I can see that the generator isn't able to deliver the fault current anyway and that the method of protection would be the generator stalling, is this correct? It seems kind of crazy to me?

I've looked into methods of calculating the max PSCC the alternator can deliver so I can see what the actiual figure is:

This requires a figure for generator subtransient reactance but I don't understand where to get this figure from. I probably don't have a full handle on the detail here so hoping someone more knowledgable can explain this to me and what the general rules are for signing off small generators under 7909 where you're not making the required regs via testing at Ze for your feeded ADS. I've got a copy of the James Eade temporary power systems and it doesn't go into detail on this specfic scenario.

I suspect one of the issues I have is a lack of understanding about the subject. I feel like some more training is needed....they don't teach you half this stuff on a basic 7909 course!!! If anyone who knows about this indepth can hekp to explain it would be really appreciated.

Cheers

It needs to be calculated as far as i can find. Google subtransient reactance and the same guy who put this video up has one for this.

 

[pc1966]

The issue in my mind is that in the event of a short I'm not making the required fire protection on the 63/3 feeder.

In most cases if the supply breaker is going to provide overload protection for the cable then it will be within the adiabatic limits of the cable.

If the PSCC is massive, for example close to a big substation transformer, you might have to worry and check the let-through energy but that is unusual. For domestic-like settings Table B7 of the OSG has some guidance limits.

If you go to a bigger example, say Hager 'x160' series MCCB at 63A nominal and PSCC of 40kA then the plot of let-through has it limiting at around 800k A2s so for usual adiabatic limit on cross section is:
S = sqrt(I2t) / k = sqrt(800,000) / 143 = 6.3mm

You would rarely be using 63A protection for a 6mm cable anyway, so it is going to be safe for typical 10mm or 16mm copper flex.

Even doing the same for steel armour, the results are:
S = sqrt(I2t) / k = sqrt(800,000) / 51 = 18mm

That would be met by 4mm 2C Prysmain cable (19mm armour), again smaller than you might be trying to sensibly use anyway.

 

[Rockingit]

The ELR needs setting to 30mA for sockets of 32A and 63A.

 

[Rockingit]

And your post very neatly examples my long-term gripe with the facade of BS7909 'qualification' - it's no such thing despite as certain person having made himself a decent business peddling it as such!

 

[EDoubleD]

I would definitely agree that 7909 is not a qualification but we all have to learn somehow and none of the domestic/commerical courses I've done touch on events in a meaningful way. Where would you suggest going for training? I assume you're reffering to James Eade?

Anyway.....what reg is that? Sure 32a and under requires 30ma. They're on 30ma RCBO but 63/3 is a distribution circuit? 30ma would be usesless given that the standing residual current should not cause operation of the RCD in normal usage. 7909 reccomends 25% if I'm not mistaken? Allowing for 1ma per appliance that's about 6 appliances on a distribution circuit? Does 7671 not state that "Supply cables to stands etc are required to have time delayed RCD's rated at 300ma". The ELR doesn't go below 300ma. Don't get me wrong I stand to be corrected by people more qualified but 100ma+ is pretty standard on all event distros on 63a+.

 

[davesparks]

To put it simply the generator cannot provide enough current to operate and mcb as fault protection.
To deal with this we make use of RCDs of various types to provide fault protection.

In you case the variable RCD at the generator will be providing the fault protection for the 63A circuit.

You don't really have a Ze because Ze is the impedance (Z) of the external (e) fault path and there is no external fault path as the generator is part of your installation.

Another common mistake is to assume that the generator is a TT supply because it has an earth rod whereas it is a TNS supply and yne earth rod is the earth reference for the star point.