Very long run of LV Cabling

Provided that the phase and neutral conductors are correctly sized WRT the fuse or circuit breaker, then there is no question of “cooking” the conductors under fault conditions.

For example, a 95mm sub main protected by a 63 amp MCB can NEVER suffer thermal damage during a fault. A modest overcurrent to say 70 amps might persist for days, but is within the long term current rating of the cable.

Likewise a 4mm sub circuit on a 32 amp MCB cant be damaged by overcurrent  provided that the full rating of the cable is available. If the fault current is low, then the fault current is reduced and the MCB will take longer to trip. But the cable does not “know” the difference between this and someone plugging in numerous portable heaters. 

 If de-rating for grouping or thermal insulation is required then I would either go up a cable size to 6mm, or go down to 20 amp MCB.

An earth fault is potentially more concerning since total reliance is placed on the RCD functioning. I would prefer a time delayed RCD at the origin and a quicker acting one at the load end.

well 95mm2 will carry the fat end of 400A /core all day at pretty much full chat. 2cores in parallel ,more like 700-800A. 

Adiabatically it will take about 20kA for a second, for an I2t limit.

What exactly will be overloaded on a 63A cct ? Even with  a massive overload the cable will run cool, just out of spec at voltage drops.

Not sure what you think will cook

Mike.

per core 95m of 95mm is about 18 milliohms, so 950m would be 0.18 ohms.  2 in parallel half this, but double it again, as the current flows there and back.

Hi Mike, thanks for the reply. It was more the final circuits after the sub main, ie a 6mm on a 32A MCB If the fault current was so low due to the long cable run that the breaker didn’t operate and carried the fault for an indefinite amount of time. but I suppose the chances of this are so small. To be honest, i suppose by the time the fault was this small from cable resistance, the voltage drop would also be an issue.

Thanks 

Thanks Broadgage, this answer makes complete sense and yes Point taken on the panel heater example.

just an FYI, all final circuits will be protected by instant 30ma RCDs

thanks again 

If 6mm cable is connected to a 32 amp MCB, then unless the cable is hugely de-rated by grouping, thermal insulation, or high ambient temperature, then there is NO  question of thermal damage from low fault currents.

At say 35 amps, the MCB might not open for days, if ever, but the cable can stand 35 amps forever.

At say 100 amps the magnetic trip will never operate, but the thermal trip will operate after a few minutes, before the cable reaches a damaging temperature.

Or put another way, a published cable rating of say 37 amps does not mean “dangerous at 38 amps”  it means “safe if protected by a 37 amp MCB” if you can find one, or in practice by any smaller size such as 32 amps.

This does not sound like a very economic solution to this supply. You need 20 kVA and you are going to spend how much on this cable? I am going to suggest that you will pay around £70 / m for around 2km of cable, that's £140, 000. Plus the cost of burying it, perhaps another £5-20k, dependant on the degree of reinstatement. You will get a new DNO supply for a lot less than that, with a local transformer, and pretty much as much power as you want. 

This is an interesting question because it shows the problem with CAD packages. Amtech will happily calculate a ridiculous answer for one, because it does not think, based on an assumed voltage drop that is really too small. You need a supply of about 32A per phase, and the cable design would allow about 400A per phase quite satisfactorily (about 250 kVA), although the volt drop would be rather high, actually extreme. The problem is volt drop, the answer is higher voltage, say 11kV, but Amtech doesn't know this. The Garbage in, Garbage out of computers is well demonstrated!

Hi Dave,

The reason I’m asking this, is because I’m doing an options report from the client.

4 Options:

850M of private cabling

20KVA Genset 

200m overhead line (private)

2Km of HV DNO cabling.

im obviously leaning towards the Genset, and when the sites more developed take it from a more local DNO supply, but I wanted to make sure option 1 was valid.

cheers 

It was more the final circuits after the sub main, ie a 6mm on a 32A MCB If the fault current was so low due to the long cable run that the breaker didn’t operate and carried the fault for an indefinite amount of time.

If you're within normal voltage drop limits, then the L-N loop impedance must also be within normal limits despite the long cable length (the massive c.s.a. making up for the long lengths) - and therefore a L-N fault (of negligible impedance) will cause a normal large fault current to flow and the MCB will open within normal time.

   - Andy.

As regards the economics of the installation, a very over-sized cable though expensive initialy has merits of simplicity and being almost everlasting.

If 3 phase is available at the supply end then I would be inclined to use a 3 phase sub main.

A generator MIGHT be attractive but needs maintenance and ideally duplication in case of failures. Running a 20 KVA generator for a load of a few hundred watts makes no sense, so a battery and inverter system is needed for low loads, all entirely doable but the cost and complication is increased.

On site PV with battery storage might be worth considering, perhaps with a generator for heavy loads. Wind power can be worthwhile if the site is windy.