Maximum distance for LV reticulation

[AusLee]

Hi,

I have an LV load that is 1,200 meters away in underground cable distance from the point of supply. It is a 3 phase LV load with 30A on the worst loaded phase.

There is an RCD protection at the origin of the cable at the point of supply, to avoid running an earth cable and to mitigate earth faults, and we've sized the cable based on voltage drop considerations.

Someone is saying that you cannot go more than 800m in LV distribution, it doesn't matter what size cable you put. Can anyone please confirm this or otherwise? is there a limit on how far your can go, provided we are willing to pay for the cable? How far have you installed an LV cable before?

Thanks.

 

[Marmite]

Providing you meet voltage drop and earth loop impedance requirements there isn't a physical limit unless you reach the point where the cable is so big it's not even manufactured. The economic limit would be exceeded way before the physical limit.
Regards
Marmite

 

[AusLee]

Hi Marmite,

Thanks for the note on the cable.

On the earth loop impedance:
1. If we have an RCD switch at the origin of the cable, would we still have to worry about earth loop impedance?

2. Come to think about it, will the RCD function properly if the cable's neutral is connected to earth? i.e. due to cable length and to save copper, instead of running 3 phase + neutral + earth we only run 3 phase + neutral, the neutral being earthed at both ends. Any particular concerns regarding RCD?

Regards.

 

[DRWeig]

I would very much like to know where the 800m limit came from, I've never heard it before and can't imagine why. Has the person saying it given you a reason? Is it perhaps in a code somewhere?

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[waross]

RCD probably won't work dependably.
The neutral may share ground fault currents.
The ground conduction may share neutral currents. No easy way to discriminate between neutral current and ground fault current.
But I may be wrong. I'm thinking about it. Skoggs??
Note currents may divide at either grounding electrode.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[AusLee]

DRWeig: that person has a "rule of thumbs book" which tried to question before but went wrong on many occasions so now taking whatever that book says for unquestionable.

Waross: I tried to work it out this way but I'm sure I' wrong. Say the contractor achieved an earthing resistance of 5 ohms at either end of the cable. And say that due to load imbalance there is a current of 20A in the neutral. And finally say the neutral conductor has an impedance of 2 ohms over its length.

Under normal conditions, the current returning from the remote load from the will find an impedance of 5 ohms (the local earth) in parallel with an impedance of 2 ohms, which is the neutral conductor, in turn in series with a parallel of 5 ohms (the earth at the source) and say o.5 ohms impedance from the neutral at the source to the neutral at the point of supply.

5 // 0.5 = 0.5, in series with 2 = 2.5 ohms
This means the 20A current will split at the load as follows:
Back to source: 20 x 5 / (5 + 2.5) = 14 A
and 6A in the local earth. That is much greater than a 30mA or even a 100mA RCD, but the figure seems too large so I must have got the circuit wrong.

 

[DRWeig]

Thanks AusLee.

I wondered aloud last night if the distance limit has anything to do with capacitive coupling? I have no idea, I've never tried to calculate same.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[waross]

That much cable will cost a lot. Can you do a cheaper installation by using a higher voltage and a set of transformers? That will also avoid current sharing issues.

The impedance of the neutral may not be as calculated due to reactance, and the ground resistance may be higher. Still, your figures illustrate that you don't need much current sharing on the ground path to interfere with the RCD operation. Of more importance, if some way is found to live with neutral currents, then the same mitigation may make the RCD insensitive to true ground faults.

That rule of thumb may be based on cost. Just thinking??
I had a similar installation. Light load and long distance. Fortunately no GFI or RCD protection and so no issues.
We had about half of the load halfway out and so that would have meant three transformers and that skewed to cost comparison. But, we calculated that by taking the voltage from 120 Volts to 480 Volts we could increase the capacity from 15 Amps to about 100 Amps on the same conductors.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[healyx]

Auslee,

So at the remote location do you have a MEN link and earth electrode? I think you'll end up getting too much earth leakage back to the source location via the mass of earth. This will cause the RCD to trip. This will be difficult to determine analytically because you'd need to know the impedance of the soil path. I don't think what you propose will work unless you could perfectly balance the three phase load at the remote site so there is no neutral current. Even the reactance of the cable over that distance will cause current imbalance.

You might need to run an earth cable and not earth the remote site. If you are feeding 3 phase motor loads only, you might not need a neutral.

Also bear in mind that an RCD will solve disconnect times for earth faults but won't help with other faults. For a small load and small cable over such a massive distance, all your fault currents are going to be very low and none of them may cause the breaker to trip in the instantaneous region. You might need to use an adjustable MCCB with instant or short-time pick-up turned all the way down.

Can you give us more specifics? Is this circuit fed from a source close to the transformer?

DRWeig has a good point about capacitance, I know this can be a big problem with long controls cables.

I have never heard of a hard limit of 800m.