Wiring up an Eddi

[Crofter]

So I'm going ahead with my plan for a ground mount array and a diverter. 

 

What's best practise for placement of breakers, isolators, etc? 

 

Thoughts on SWA vs a duct for the DC run? 

 

The inverter in using is IP rated but I think it's best to plan for it to be in the house anyway? It's about a 25m run from the array.

[Crofter]

So I've been told that it's best to place the inverter up at the array. I think I'll need to make some sort of enclosure for it due to the weather we get!

 

I'm also a bit confused about cable sizes. By my understanding, if I have say ten panels in series, I'll have 300v minimum. For a 3kw array, that means about 10A. So I should be OK with a 1.5mm2 cable over my 25m run. But everything I read says I must use 4mm2 minimum. Am I fundamentally misunderstanding something here?

[JohnMo]

Run long cables in DC, so they take any voltage drop on that side, (less of an issue for the inverter - doing on the AC side cause high AC voltage and it could result in the inverter tripping.

 

When I did my install PV rated SWA didn't exist, it now does. I ran my cable on the surface - easy to see and less likely to get accidentally damaged.

 

10 minutes ago, Crofter said:

about 10A

New panels are generally maxing out at nearer 14A. The panels will generate full voltage even in low light, the amps increases with more irradiance.

 

Find a DC voltage drop calculation online, I used I thing 6mm² for about 40m.

[MikeSharp01]

16 minutes ago, Crofter said:

So I should be OK with a 1.5mm2 cable over my 25m run

Far too small - 4mm is minimum and 6mm prob better for run lengths.

[ProDave]

for the DC runs, easiest to buy proper solar PV DC cable, sold as singles usually 4mm or 6mm and run them in conduit.

 

Probably best to have a DC isolator switch next to the panels, and another where the DC cables go into the inverter.

[SteamyTea]

Yes, do the long runs on the DC side.

Ideally you want DC isolators at the panel end, and then depending on inverter, maybe on nearer for convenience. A lot of new inverters have DC isolators built in.

An inverter can generate a lot of hot air when running at 100% plus, so if indoors, make sure there is adequate ventilation. AC isolator is usually near the inverter and are nearly always used to isolate before the DC (takes the load of the inverter.

As for cabling size, go for minimum of 6mm (even if the panels have 4mm tails). The DC losses (below 2% ideally) will only be a problem when you are generating close to maximum, most of the time you will only be at a small fraction of peak amps.

Does depend how long the run is though.

[Crofter]

Thanks everyone.

I'd initially read/heard that the long run should be before the inverter. Then somebody (on Faceache) said it was the other way round.

Does the choice of panels affect this decision?

E.g. if I had ten 300w panels, I'd be running about 300v DC or more. But if I had seven 500w panels instead, I'd have a lower voltage.

You'd think you'd want the longest run to be where you run the highest voltage, for lowest losses and smaller cable sizes.

 

Bear in mind I'm coming to this with fairly extensive off-grid experience so I have a certain amount of intuition which isn't necessarily helpful

[JohnMo]

Just now, Crofter said:

Does the choice of panels affect this decision?

No. Say your inverter trips out at 253V, grid on that day is already at 250V. On the AC side the inverter, it pushes out grid volts, plus a bit, otherwise it cannot break into the grid. If you take all the voltage drops on the AC side the inverter adds volts to make sure it can break into grid, plus it adds the voltage drop. So on a high voltage day your inverter may trip off. It does care what volts it gets on DC side 

[Crofter]

9 minutes ago, JohnMo said:

No. Say your inverter trips out at 253V, grid on that day is already at 250V. On the AC side the inverter, it pushes out grid volts, plus a bit, otherwise it cannot break into the grid. If you take all the voltage drops on the AC side the inverter adds volts to make sure it can break into grid, plus it adds the voltage drop. So on a high voltage day your inverter may trip off. It does care what volts it gets on DC side 

That makes sense, thanks.

 

I'm still wondering why the DC cables need to be so beefy.

I did my initial calculations based on twelve 285w panels. That gives a minimum system voltage of 348v, and a maximum current of about 10A.

Even a skinny little 1.5mm2 cable will only show a voltage drop of 1.82% over 25m.

Upgrading to 4mm2 gets me down to 0.6%, and 6mm2 would be 0.45%. I just didn't think that was necessary?

 

I could use a smaller number of larger panels, e.g. seven, giving me 234v/16A (gives a slightly smaller array overall). In this case 1.5mm2 is definitely too small (more than a 4% drop) but 2.5mm2 looks pretty good to me (2.6% drop).

4mm2 gives 1.6% drop.

6mm2 gives me a 1% drop.

 

I've built a lot of off grid stuff where this kind of voltage drop is of no concern. but I appreciate that things are a bit different. When you have a panel pumping out 17v and you want 14v to charge a battery, you can afford to lose a whole volt somewhere and it will still work.

 

I'll happily stump up for the chunkier cable, I just like to know why...

 

 

[SteamyTea]

44 minutes ago, Crofter said:

I did my initial calculations based on twelve 285w panels. That gives a minimum system voltage of 348v, and a maximum current of about 10A.

PV modules (the individual cells) are fixed voltage devices, basically diodes), as @JohnMo mentioned earlier, it is the current that varies with light intensity.

Modules are tested to a standard format, "The reference condition called standard test conditions (STC) is commonly used and assumes 1000 W/m² solar irradiance, AM1.5 spectrum, and a cell temperature of 77°F(25°C)" (more here), it is not difficult to exceed those conditions (even in the UK), so a lot easier to put in a safety factor.

Putting cables into conduit can affect the amount of current they can carry (similar to passing cables though insulation).  There are standards for this as well.

DC cables also carry, for a given size, a different amount to AC cables (usually more but not much in this situation, more at higher voltages).

On this site, there is no price difference between them.

 

[Beelbeebub]

Chunkier DC cables gives you headroom should you upgrade to more/better panels in the future. 

 

The resistance of 6mm cables, about 2/3 that of 4mm2 (and 1/4 of the 1.5mm cable) is lower which will reduce your i2r losses (not as much as lowering your current but there isn't much you can do about that) 

 

There are DC specific armoured cables now (the common AC ones are not suitible for DC) which makes doing the runs much neater and easier. 

 

 

[Crofter]

20 minutes ago, Beelbeebub said:

Chunkier DC cables gives you headroom should you upgrade to more/better panels in the future. 

I appreciate that, I just thought maybe I was missing something.

I'm not sure how likely an upgrade would be. I'll be maxing out my permitted development with this ground mount array, and the next sensible move would be adding something to the roof.

I also have the other property which would benefit from a similar setup, and I'd likely do that project before I did any upgrades to this system.

 

20 minutes ago, Beelbeebub said:

 

The resistance of 6mm cables, about 2/3 that of 4mm2 (and 1/4 of the 1.5mm cable) is lower which will reduce your i2r losses (not as much as lowering your current but there isn't much you can do about that) 

 

There are DC specific armoured cables now (the common AC ones are not suitible for DC) which makes doing the runs much neater and easier. 

 

 

I hadn't heard that standard SWA wasn't suitable for DC. I thought that so long as you stayed within the voltage rating you'd be good.

Would something like this be suitable?

https://www.superlecdirect.com/6942x-6mm-2core-bs5467-xlpe-swa-pvc-cable-harmonised-black/

 

(the datasheet for the above cable shows DC and AC values. The 6mm2 is rated to 53A when buried in the ground, which is a lot of headroom for a system that should be generating around 15A)

[Beelbeebub]

I think it's that the regs require the conductors to be double insulated within the armour. 

 

Ie you take double insulated conductors, then pack them into an armoured sheath, then wrap that in insulation (which is mainly to protect the armour) 

 

Most SWA is single insulated inside the armour ie each cable is single insulated, 

 

Ie you take single insulated conductors (like you woikd find in a twin and earth) then wrap those in the armour. 

 

I think it's because DC shorts between conductors are much worse than AC for the same voltage and current. So the individual conductors need that extra level of protection. 

[Beelbeebub]

https://www.doncastercables.com/cables/20/89/PV-Ultra--/PV-Ultra-----Double-Insulated-Multicore-DC-Cable/

 

The do 4 core so you can have 2 strings if required. 

 

Regarding upgrades it depends a bit on which panels you're putting in. 

 

Current "mid range" panels are in the 450w range with some premium panels going over 500w, a few years go the 450w panels were premium and most were 400w or lower 

 

In 10 years time the cheapest panels might be 550w and the premium ones 650w 

 

So if you reolwce/upgrade your current will go up even if your array stays nominally the same. 

 

Also we don't know what planning will do in the future, it may be they change the permitted development allowance and you could add more panels. Be nice if the cables were already sized for that, especially if relaying them would be a faff - if the route would be a doddle to add an extra cable don't worry too much. But if it"s a sod, just do it once 

[Crofter]

16 minutes ago, Beelbeebub said:

https://www.doncastercables.com/cables/20/89/PV-Ultra--/PV-Ultra-----Double-Insulated-Multicore-DC-Cable/

 

The do 4 core so you can have 2 strings if required. 

 

Regarding upgrades it depends a bit on which panels you're putting in. 

 

Current "mid range" panels are in the 450w range with some premium panels going over 500w, a few years go the 450w panels were premium and most were 400w or lower 

 

In 10 years time the cheapest panels might be 550w and the premium ones 650w 

 

So if you reolwce/upgrade your current will go up even if your array stays nominally the same. 

 

Also we don't know what planning will do in the future, it may be they change the permitted development allowance and you could add more panels. Be nice if the cables were already sized for that, especially if relaying them would be a faff - if the route would be a doddle to add an extra cable don't worry too much. But if it"s a sod, just do it once 

It's not a hard run- just across some rough ground at the back of the garden. I'm not digging up a driveway, or even a lawn. So I don't think I'll regret it too much if I do end up having to re-do it.

 

Anyway... cabling aside, with the rest of the installation, what would be best practise:

- breaker and/or isolator up at the panels?

- breaker between the Eddi and the immersion?

- anything else that might not be obvious to me?

[SteamyTea]

12 minutes ago, Crofter said:

breaker and/or isolator up at the panels

Definitely, you don't want a live conductor, AC or DC, what you cannot isolate at the source end.  Make sure that isolator can handle the DC current.

 

14 minutes ago, Crofter said:

breaker between the Eddi and the immersion

Yes.  I am not sure if there should be one before it.  Probably does need something to safely isolate it, but the MCB/RCD/RCBO may well do that.

What do the regulations say?

[cjsparkey]

MCB/RCD/RCBO To DP isolator for Eddi, load side of the Eddi to immersion via a local isolator.

Technically the local isolator for the immersion might not be needed if the Eddi is local to the immersion but it’s handy if you ever need to isolate the immersion without needing to shut down the Eddi.