Changes to Solar PV Strings

[grahamsmythe]

Some time ago, we had a PV system installed, and ever since, I've been unhappy with the way it's been wired, so would appreciate any thoughts that you may have.

 

All DC wiring is 4mm.

 

As you will see from the image below, we have two strings of ten panels, and these are connected together via a DC isolator in the loft.

 

When the cable exits the loft isolator, both positive and negative go down into the garage via four 4mm cables.

 

When the cables reach the garage, they go through another DC isolator before exiting through a single cable and then into the inverter.

 

Currently all works fine.

 

However, our inverter has two sets of inputs to facilitate two strings, so what I'd really like to do is take each string down to the garage separately, so we're able to connect the two strings separately into the inverter, so we can see how the strings are performing individually.

 

 

The following diagram shows how I think it could be done using the existing wiring and isolators.

 

 

As stated previously, I'd appreciate any thoughts that you may have.

 

Edited November 2, 2019 by grahamsmythe

[SteamyTea]

My initial thoughts is that if all the modules are wired in series the voltage will go over the LV limit of 1000V.  This does depend on which modules are fitted.

What inverter do you have fitted?

[grahamsmythe]

The panels are 195w (20x195w = 3.9kw).

 

The inverter is a Diehl AKO Patinum 4300S http://www.raine-or-shine.com/dloads/PV/Inverters/Diehl-AKO-Platinum-Inverter-Manual.pdf

 

 

[SteamyTea]

The 4300S has a maximum DC voltage of 780V.

Do you know what voltage your modules are.  Do you have a spec sheet for them?

Does look like it would be better off wired as two strings.

 

Have you had any errors showing on the display?

Edited November 2, 2019 by SteamyTea

[Jeremy Harris]

19 minutes ago, SteamyTea said:

My initial thoughts is that if all the modules are wired in series the voltage will go over the LV limit of 1000V.  This does depend on which modules are fitted.

What inverter do you have fitted?

 

 

Looks like the two strings are wired in parallel to me, so OK in terms of max voltage.

 

The usual reason for wiring two strings to two separate MPPT inputs on the inverter is if the strings may produce slightly different outputs, from being differently aligned, or if the strings are of unequal number of panels (for example our system has12 panels in one string and 13 panels in the other).

 

In this case, I doubt there would be the slightest difference if the two equal strings were wired to separate MPPTs.  There might be a miniscule difference from the very slightly reduced voltage drop in the separate 4mm² DC wires, but not enough to be able to detect, in terms of performance, I suspect.

 

No reason not to do it if you wish, though, especially as it looks as if you can isolate the panels to make the downstream wiring safe to work on.

[SteamyTea]

Yes, it does seem the voltage is OK, maybe a problem with Sanyo HIT modules as they are 55V max load modules.

 

Must be a large DC isolator.  Had one burn out once when the electrician miss wired one.

[grahamsmythe]

This is the specification of the panels…

 

[grahamsmythe]

So, just to be clear, am I right in thinking:

 

1. Switch the DC isolator in the loft to off.

2. Switch the DC isolator in the garage to off.

3. Isolate the AC isolator in the garage to off.

4. Unplug the wire that goes from the loft isolator to the panel (x4).

 

I'm then safe to alter the wiring as described previously?

[SteamyTea]

Switch the AC to the inverter off first, that takes the load off the DC side, then turn the DC isolators off.

Or do the work after 5 PM.

[grahamsmythe]

Is it safer to do all the connections etc at night? ?

[Jeremy Harris]

18 minutes ago, grahamsmythe said:

Is it safer to do all the connections etc at night? ?

 

 

Slightly, but if you turn off the AC, then turn off the DC at the isolator nearest the panels, then do a dead check, you're safe enough.  Usual golden rule applies - never, ever, assume a switch works, always test to be sure it has!  Second golden rule - never use a voltage detector pen/screwdriver, they are notoriously unreliable.

[Ed Davies]

On 02/11/2019 at 17:04, Jeremy Harris said:

Slightly, but if you turn off the AC, then turn off the DC at the isolator nearest the panels, then do a dead check, you're safe enough. 

 

It looks to me like @grahamsmythe will be changing the wiring on the DC isolator nearest the panels. Dunno about anybody else but I think I'd do that part, at least, at night.

 

Why does the original wiring use two contacts to break each line? There's no possibility that they need the extra distance to fully break any arc, is there?

 

An aside: I've just looked at my three multimeters for suitability for this job. The two cheapo ones with CAT classifications of some sort only go up to 300 and 600 volts. The better one which does go up to 1000 volts and which I'd probably distrust a bit less doesn't actually have a CAT classification and has lots of bare metal on the probes. Not having a dedicated tester, in practice for something like this I'd probably borrow the better protected probes off one of the CAT classified meters to test dead on the DC side for any work done while there could be any glimmer of sunlight. It's probably not that big a deal but the sort of thing which is worth thinking about. Shorting the panels by touching the wrong bits of metal with a meter probe isn't likely to be a big problem - it's only ~10  A max - but the arc you could draw unshorting them could be illuminating.

[ProDave]

It is normal to use 2 poles for DC isolation. Almost certainly for arc quenching.

 

connecting the 2 strings independantly won't do any harm, and might give a slight improvement if any panels get any shading, so I would do it.  And I would probably do it on a dark night like tonight.

[grahamsmythe]

Thanks gents. Much appreciated.

 

I did change the wiring on the isolator in the loft, but as mentioned by @ProDave I now only have one isolator for DC isolation, so I'll get another DC isolator to do it properly.

 

My multi-meter goes up to 1000v DC so was able to prove it was dead. However, I did take the precaution of doing the work at night!

[SteamyTea]

45 minutes ago, grahamsmythe said:

I did take the precaution of doing the work at night!

Can you work out if you are getting anymore power out of the system than before?

[grahamsmythe]

On 05/11/2019 at 12:17, SteamyTea said:

Can you work out if you are getting anymore power out of the system than before?

Yes, I'll try to do that once I've finished it.