Depth for buried DC cables

[Beelbeebub]

Does anyone have a reference for the depth that DC cable needs to be buried - I assume the same as armoured AC (0.5m in general) but wanted to double check.

 

I was going to use the specific armoured stuff 

[JohnMo]

If you don't really need to bury l wouldn't - would surface run, then you can see them.

[Beelbeebub]

3 hours ago, JohnMo said:

If you don't really need to bury l wouldn't - would surface run, then you can see them.

For at least some of the run they will go across an agricultural field (sheep) that is occasionally flail mown and then also across a lawn which is regularly mown and trafficked.  Also across a tarmac drive.  It's (hopefully) a single ground mounted array that serves several houses.

 

The other alternative is overhead poles but that has issues too. 

[JohnMo]

You need to bury below deep plow depth. Although sheep today, you have no idea what will occur in year. So that is 910mm to 1000mm

[Beelbeebub]

9 hours ago, JohnMo said:

You need to bury below deep plow depth. Although sheep today, you have no idea what will occur in year. So that is 910mm to 1000mm

Good point.

 

Though I can guarantee the field isn't going to get ploughed anytime due to the slope and it also being an orchard.

 

It's a short run from the array site to the edge of the field so I could mount it above ground on some timber rails.  If I run it behind the hedge in a similar manner I could minimise the distance underground as @JohnMo suggests.

 

Looking at the site I realise there already is a telephone line overhead along the fence.

 

I wonder if I could run the cables overhead for a good chunk.

[Dillsue]

If it's an armoured cable and you have a route along a fence, why not clip it to the fence rails or tywrap it to a strand of fencing wire.

[Beelbeebub]

8 hours ago, Dillsue said:

If it's an armoured cable and you have a route along a fence, why not clip it to the fence rails or tywrap it to a strand of fencing wire.

I don't have a fence only a hedge, and at one point I have to pass a gate and another a drive way.  In the field I could see effectively building a fence to carry the cable - i'll probably do that.  at least the buried sections will then not be in a field.  the biggest bugger is there are a ton of other services around, gas, mains, water and sewage plus some telephone.

 

Potentially I could go overhead the whole way, it's just passing over the drive way and gate are a bit untidy.

[Nickfromwales]

For this distance, sending DC back is nuts. Abandon this is my 2 cents, as you'll need huge CCA of the cables.

 

Put the inverter out with the array and send AC back; same we did a while back on a ground mount some 135m away from the house, where the client wanted the array in a meadow vs the untouchable roof of his listed home.

[Beelbeebub]

5 hours ago, Nickfromwales said:

For this distance, sending DC back is nuts. Abandon this is my 2 cents, as you'll need huge CCA of the cables.

 

Put the inverter out with the array and send AC back; same we did a while back on a ground mount some 135m away from the house, where the client wanted the array in a meadow vs the untouchable roof of his listed home.

Not a bad idea, it would definitely cut cable costs but I think i'd I be giving up the power cut backup function?

 

And is there a problem with exporting the earth? The panels will all be earthed together by virtue of being fixed to a metal frame stuck to the ground.  But the houses they supply are 50 to 100m away. 

 

wouldn't the same issue with car chargers and remote earth/faults raise it's head?

[Roger440]

4 minutes ago, Beelbeebub said:

Not a bad idea, it would definitely cut cable costs but I think i'd I be giving up the power cut backup function?

 

And is there a problem with exporting the earth? The panels will all be earthed together by virtue of being fixed to a metal frame stuck to the ground.  But the houses they supply are 50 to 100m away. 

 

wouldn't the same issue with car chargers and remote earth/faults raise it's head?

 

Im doing exactly what nick suggested. Mines 130m away. Trencher is booked!

 

What is is your thinking around the power cut back up function. Assuming you will have a battery?

[JohnMo]

Doing long distance AC can lead to inverter switching off depending on the voltage of the grid at the house(s). Taking a voltage drop on the DC has zero effect.

 

When doing the AC over the distance the inverter has to produce more volts to overcome the AC line voltage drop. 

 

One of my arrays has 80m of DC cable zero issues and cable isn't that large. If doing DC make sure the voltage is high, low voltage DC for a given amps is big, high voltage for a same amps gets smaller

[-rick-]

6 hours ago, Nickfromwales said:

For this distance, sending DC back is nuts. Abandon this is my 2 cents, as you'll need huge CCA of the cables.

 

Put the inverter out with the array and send AC back; same we did a while back on a ground mount some 135m away from the house, where the client wanted the array in a meadow vs the untouchable roof of his listed home.

 

Surely this depends on the voltage the array/string is at? With modern inverters often wanting arrays of 800-900v, you'd need less CCA on DC than AC. Even the older but common 350v arrays/strings would be better over DC than AC.

 

DC is more efficient at long distances. The conversion between different voltages and safety of the system.

Edited 3 hours ago by -rick-

[Beelbeebub]

48 minutes ago, Roger440 said:

 

Im doing exactly what nick suggested. Mines 130m away. Trencher is booked!

 

What is is your thinking around the power cut back up function. Assuming you will have a battery?

In favour hybrid systems so the battery generally needs to be next to the inverter.

 

If I use the AC method then there is one AC line (a line, neutral and earth) going to the remote inverter connecting to the meter tails between the meter and the consumer unit and I need a switchover box at the house to isolate from the mains in a power cut so I don't energise the mains.

 

I could arrange for power cut cover but there would need to be a Comms line as well which cuts onto the cost savings.

 

 

[Beelbeebub]

6 minutes ago, -rick- said:

 

Surely this depends on the voltage the array/string is at? With modern inverters often wanting arrays of 800-900v, you'd need less CCA on DC than AC. Even the older but common 350v arrays/strings would be better over DC than AC.

 

DC is more efficient at long distances. The conversion between different voltages and safety of the system.

 

48 minutes ago, JohnMo said:

Doing long distance AC can lead to inverter switching off depending on the voltage of the grid at the house(s). Taking a voltage drop on the DC has zero effect.

 

When doing the AC over the distance the inverter has to produce more volts to overcome the AC line voltage drop. 

 

One of my arrays has 80m of DC cable zero issues and cable isn't that large. If doing DC make sure the voltage is high, low voltage DC for a given amps is big, high voltage for a same amps gets smaller

Good point 

 

The array will be fairly large (maybe 8+ panels for each house) so the voltage will probably be higher than mains so the current lower.

 

Plus we have had high voltage issues in the past.

[Alan Ambrose]

I've been mulling this situation over too. I don't think it's that uncommon with e.g. (a) panels on a roof, (b) modern all-in-one stacks which tend to have all the bits (inverter, battery, gateway, car charger etc) all at the same location. And as batteries can't be an an attic these days, it suggests the stack is outside or in an outbuilding and maybe some distance away from the panels - which implies a longish DC connection. Either that, or the inverter is in the attic and the battery some distance away, which implies likely an ac connection, but then a split stack.

[Alan Ambrose]

>>> maybe 8+ panels for each house

 

So maybe 400V max (at low temperatures) & 15A max (in summer sun). 

Edited 52 minutes ago by Alan Ambrose

[-rick-]

Modern panel open circuit voltage is often ~45v so 8x45 is 360v

[Alan Ambrose]

Agreed and the temp co-efficient might push that up to 50V at low temperatures.

[Alan Ambrose]

So, for the OP @Beelbeebub that suggests 2-core 4mm PV ultra SWA (PV-ULTRA2C4.0SWA) rated at 52A and £9 p.m. 

 

Power loss at 15A is 15^2 * 5 (ohms per km) * 2 conductors i.e. about 220W per 100m (and you're carrying ~4kW at that point, so 6%). Voltage loss 15 x 5 / 10 * 2 = 15V i.e. ~4% (per 100m again). Not insignificant.

 

>>> And is there a problem with exporting the earth?

 

Good question.

Edited 15 minutes ago by Alan Ambrose

[JohnMo]

 

 

13 minutes ago, Alan Ambrose said:

Power loss at 15A is 15^2 * 5 (ohms per km) * 2 conductors

DC circuits use resistance calculations only. This makes voltage drop simple and linear: V = I × R.

 

Definitely do minimum of 6mm² any smaller you are wasting power. Then you are at or below 3% voltage drop and for a minor uplift in cost. Most panels these days come with 6mm² tails not sure why you would down grade them with 4mm².

 

 

Do multi core then one cable will carry multiple strings, so 4 core two strings etc. easier to route one cable instead of loads.

[-rick-]

8 minutes ago, Alan Ambrose said:

>>> And is there a problem with exporting the earth?

 

No an electrician but I thought over the distances we are talking about you need to install a local earth rod. Earthing would only be for the metal work and wouldn't touch the DC AFAIK.