Long distance PV set up

[jayc89]

One of my many projects is our PV set up. Eventually I plan on having 3x arrays;

 

- Shed roof (essentially ground mounted) approx. 5.5kw

- Garage roof approx. 3kw

- House roof approx. 3.5kw

 

Eventually there will also be some battery storage too, as I don't plan to feed any back to the grid, either in the garage (so could store power from both garage or shed) or in a lean-to near the meter box (so could store from all 3x arrays). If the former, I might never install the house roof array.

 

The problem I'm currently tackling is the distance between the "shed" and garage to the meter box. The garage is approx. 30m and the shed is a further approx. 65m behind the garage, so the longest run would be a max of 95m. In order to keep the voltage drop below 1%, assuming an AC cable of course, I'm calculating that would need a 50mm2 cable, which would be ludicrously expensive. Ideally I'd only have one cable coming from the garage back to the meter box which is carrying the supply from both the "shed" and the garage arrays. 

 

Would would you suggest is the most efficient solution here, both in terms of cost and performance?

 

 

[ProDave]

Do the long run as DC cable, with the inverter very much closer to the meter box.

 

As you rightly say, voltage drop (rise) on the ac side will lead to the inverter tripping.  Voltage drop on the dc side will just be a tiny % of power lost as the inverter is very flexible about input voltages.

[jayc89]

20 minutes ago, ProDave said:

Do the long run as DC cable, with the inverter very much closer to the meter box.

 

As you rightly say, voltage drop (rise) on the ac side will lead to the inverter tripping.  Voltage drop on the dc side will just be a tiny % of power lost as the inverter is very flexible about input voltages.

 

Makes sense - so I could use a 6mm2 SWA from the arrays back to the lean-to near the meter, and house the inverter + any batteries there? Then the AC run is a max of 2m into the meter box.

[JohnMo]

I did my long runs in DC. Keep amps low otherwise cables start to get huge very quickly. link to the threads below.  It works a treat, no issues at all so far.

 

 

 

 

[Bramco]

We also did ours in DC - about 50m to the top of the field.

We have 12 panels in all - 6.5kWp.   The cable was a 4 core cable, so there are 2 groups of 6 feeding the 2 DC inputs on the inverter.  

The ground mounts are key clamp system with 16 1m poles whacked in the ground as the support.

We also set the panels at about 45degs because at the optimum angle (30deg ish) we'd get massive summer output and low winter output.  45degs gives us lower summer output and higher winter output.  We're just about getting the maximum out of them about now.

 

Simon

[jayc89]

Trying to come up with a sane design for these runs. All will be in serial strings, as parallel will result in too high amperage (I think).

 

"shed" - 10x 550w panels @ 49v each. 

garage - 5x 550w panels @ 49v each.

house roof - 6x 550w panels @ 49v each.

 

AFAIK, when running serial strings, voltage is accumulative (volts x panels) but the wattage remains constant (550w in my case). That being the case, to stay within a voltage drop of 1%, the cable runs required seem to be;

 

"shed" - 10x 49v = 490v @ 550w = 1.12a over 95m distance, requires 0.8917mm2 cable (1.5mm2)

garage - 5x 49v = 245v @ 550w = 2.24a over a 30m distance, requires 1.1264mm2 cable (1.5mm2)

house = 6x 49v = 294v @ 550w = 1.87a over 25m distance, requires 0.653mm2 cable (1.5mm2)

 

tl;dr all runs could be done in 1.5mm2 cable which seems far too low, especially as everything I've seen online suggests between 4 and 6mm2 SWA (but that is generally when talking about AC).

 

Am I mis-calculating something?

 

[JohnMo]

Something well a miss. Your panel amps stays the same, if you have an 10 amp panel , you have 10 amps, you need to calculate voltage drop at volts x number of panels and panel amps, ignore the watts.

 

 

 

 

 

 

[Bornagain]

10 minutes ago, jayc89 said:

Trying to come up with a sane design for these runs. All will be in serial strings, as parallel will result in too high amperage (I think).

 

"shed" - 10x 550w panels @ 49v each. 

garage - 5x 550w panels @ 49v each.

house roof - 6x 550w panels @ 49v each.

 

AFAIK, when running serial strings, voltage is accumulative (volts x panels) but the wattage remains constant (550w in my case). That being the case, to stay within a voltage drop of 1%, the cable runs required seem to be;

 

"shed" - 10x 49v = 490v @ 550w = 1.12a over 95m distance, requires 0.8917mm2 cable (1.5mm2)

garage - 5x 49v = 245v @ 550w = 2.24a over a 30m distance, requires 1.1264mm2 cable (1.5mm2)

house = 6x 49v = 294v @ 550w = 1.87a over 25m distance, requires 0.653mm2 cable (1.5mm2)

 

tl;dr all runs could be done in 1.5mm2 cable which seems far too low, especially as everything I've seen online suggests between 4 and 6mm2 SWA (but that is generally when talking about AC).

 

Am I mis-calculating something?

 

No, when running serial strings the Voltage goes up as you assume, the current is the same, and hence the total power goes up, so for example the shed system will have 490V and 5500 Watts, hence a current of about 11 Amps.

 

11 Amps over 95 meters will require a cable much bigger than 0.8917 mm2.

 

Remember that the cable will go there and back.

[JohnMo]

So shed would require 50mm2.

 

You may need to plan this a little differently.

[jayc89]

Ah ok, I thought only voltage increased in serial strings, whereas current increased in parallel strings... 

 

I'm still at a bit of a loss though as I've used a couple of online calcs using 490v, 11amps over 95m and they're suggesting a cable size of 10mm2, which would be doable. How are you getting 50mm2?

 

If, the shed run is going to be problematic, what about an inverter in the garage. So a 65m DC run from the shed, then a 30m run of 25mm SWA (AC) from the inverter in the garage to the meter box, which, again if my calcs are right this time, be good for up to 9kw (shed + garage arrays)?

[JohnMo]

I used the online calculator in the previously linked threads.

 

[jayc89]

 

4 minutes ago, JohnMo said:

I used the online calculator in the previously linked threads.

 

 

That was one of the ones I used, I must be doing something wrong (appreciate the cable temp is, hopefully, wrong, but reducing that only reduced the losses anyway);

 

[JohnMo]

Just went through it again, I was looking at volts drop instead of volts % drop. 10mm2.

[jayc89]

So a 10mm2 SWA would be suitable for the longest route ("shed"), which runs past the garage. Can I use a 3-way joint connector to tee into the "shed" feed and save running a 2nd cable back to the meter box from the garage?

[Dillsue]

If you "tee into the shed feed" your calcs go out of the window as you ve got 8.5 kw in the section of cable from garage to inverter. You're probably best getting a spark to spec all this for you and test it when it's done