Help with string inverter problem

Martyn1981

michaels wrote: »

Where do you get his data from? Thanks

Hiya. Just google your inverter make and model, you'll get loads of hits, the main site, retail shops etc etc, then just look for a tech sheet.

The sheet will give an efficiency figure, but also a euro efficiency rating (they'll be similar). Go with the euro rating as I think(?) it's more comparable to other inverters.

If you get stuck then just post the make and model and I'll find you a sheet. Shame Swithenbanks has gone bust as they were a real handy shortcut to all of this stuff.

BTW for any sizeable inverter, say 2.5kW up, the figure will probably be around 95% to 98%.

Mart.

PotentialEnergy

Mart
The 22 was peak. The total readings for July since my system was installed is as below :
July 2011 total 459
July 2012 total 455.9
July 2013 total 572
July 2014 total 514.2
July 2015 total 466.6
July 2016 total 242.9
Given the relatively gloomy July 2016, I think it is fair to say that the system is only performing at about 50%. But, I don't suggest to understand the number of strings etc however the totals speak for themselves.
Thanks.

Martyn1981

PotentialEnergy wrote: »

Mart
The 22 was peak. The total readings for July since my system was installed is as below :
July 2011 total 459
July 2012 total 455.9
July 2013 total 572
July 2014 total 514.2
July 2015 total 466.6
July 2016 total 242.9
Given the relatively gloomy July 2016, I think it is fair to say that the system is only performing at about 50%. But, I don't suggest to understand the number of strings etc however the totals speak for themselves.
Thanks.

Cheers I thought a daily figure (as in average) of 22kWh couldn't be right.

Yes, your figures do suggest a halving. Quick lesson on strings.

Each inverter has a minimum and maximum voltage that they can work with. For some single string inverters this can be pretty high, perhaps 700V+. But most are lower perhaps 500V. [The SMA SB 3800 is 200V to 500V, but more importantly has a MPPT range of 200V to 400V, so that's what you'd want to aim for.]

Let's imagine a fictional inverter with a range of 200V to 400V, and you want to install 16 panels, each with a voltage of 30V. The total voltage would be 16 x 30 = 480V, so too much for the inverter, so you split them into two identical strings of 8 panels, or 240V. Note this will double the amperage, so the wattage (volts x amps) is still the same.

Your inverter has up to 3 strings, so it's fair to bet 2 have been used. Your reading suggest a loss of generation from one of the strings.

Now if a competent engineer has been out, then that suggests to me that the inverter isn't working properly and is no longer processing one of the strings. But is there any chance the engineer got it wrong and one of the strings is 'broken', and the inverter isn't seeing the DC leccy?

At this point I run clean out of ability, as I have no way of knowing how you work out which is the case, sorry.

Mart.

pinnks

Silly question perhaps but has the "dog" turned off one of the DC isolators? My system uses two strings and each has a big DC switch to isolate it from the inverter. That would "break" the string in this context as much as a broken wire or defunct panel

gefnew

Hi PE
If you ring the sma England support telephone line, and explain to them that you seem to have a issue with your inverter than they may talk you through some diagnostic checks that could be displayed on the inverter to give you a better idea of what is occurring and in which direction to look. On my sb4000tl it shows on the display both strings and there outputs. may be worth getting the manual out or download from sma DE,
Have you tried a Bluetooth scan to see if the inverter shows up this normally shows as sma and some numbers, if you can see it download sunny explorer and set up it will give you vast amount of info.
Regards
Geoff

michaels

Martyn1981 wrote: »

Hiya. Just google your inverter make and model, you'll get loads of hits, the main site, retail shops etc etc, then just look for a tech sheet.

The sheet will give an efficiency figure, but also a euro efficiency rating (they'll be similar). Go with the euro rating as I think(?) it's more comparable to other inverters.

If you get stuck then just post the make and model and I'll find you a sheet. Shame Swithenbanks has gone bust as they were a real handy shortcut to all of this stuff.

BTW for any sizeable inverter, say 2.5kW up, the figure will probably be around 95% to 98%.

Mart.

When I looked up PVGIS I think I just left the efficiency at the default (90%?) - should I have looked up the invertor figure and used this?

Martyn1981

michaels wrote: »

When I looked up PVGIS I think I just left the efficiency at the default (90%?) - should I have looked up the invertor figure and used this?

Hi. Is there a specific inverter section (I just tend to go with defaults). The section on losses is set at 14%, and you could certainly change that a percentage or two. The help section even suggests this if you have a higher efficiency inverter, though I'm not sure what their default is based on.

Going back to the early year discussions on here, we all seemed to be doing too well, but PVGIS revised the figures upwards, I believe this was based on the 10yr average having improved. However, even back then some friendly installers on the Navitron site were suggesting that the 14% figure was probably a bit too high.

My personal opinion is that the estimates are a tad low, especially as I seem to do ok, despite a small amount of shading. But on the FAQ's I've been reluctant to suggest 'tampering' with the numbers so that people don't get misled on possible income, and make a bad investment decision based on 'best possible' figures. I'm not sure if that's the correct approach to take, but you can probably see the dilemma I face as the FAQ's have my name on them!

If I'm totally honest, I'm surprised the default losses haven't snuck down at least 1% over the years we've all been playing with/enjoying PVGIS.

Mart.

PotentialEnergy

The system has a dual input DC Isolation Box(labelled IP66/IP67) with just one a single on/off so as the problem of halved energy existed before the electrician came, it presumably wouldn't be this unit.

But a few points and a question :

The SB3800 Single MPPT has 3 strings and the array is configured in 2*7 so 2 of the three strings were used to avoid overloading an individual string i.e. 7*285V is within range whereas 14*285V would exceed.

If just one string has gone defective couldn't the third string be used? After all, I’m still getting 50% generation which suggests one of the 2 used strings is fine and presumably the MPPT aspect is still functioning? A 3 string device can surely provide resilience in the event of just one string going defective.

If the third string doesn't work it rather suggests some open circuit issue exists on one of the arrays and in that case the SMA inverter is actually OK

Thanks for any more guidance/views. :idea:

Martyn1981

PotentialEnergy wrote: »

If just one string has gone defective couldn't the third string be used? After all, I’m still getting 50% generation which suggests one of the 2 used strings is fine and presumably the MPPT aspect is still functioning? A 3 string device can surely provide resilience in the event of just one string going defective.

TBH that's something I've been wondering if it's simply a matter of some fault with the connection/connector where one string plugs in. But you're back to needing a good sparkie/inverter guy having a look.

PotentialEnergy wrote: »

If the third string doesn't work it rather suggests some open circuit issue exists on one of the arrays and in that case the SMA inverter is actually OK

Thanks for any more guidance/views. :idea:

That's what Z suggested a while back, and hence the need to play with isolators, but that's a problem then if they both isolate together. So, you've guessed it, you need the sparkie to see if there actually is power on both strings before the inverter.

The inverter screen should tell you the V's and A's of both strings. If one is zero, then you know it's a string fault, but not whether it's the string or inverter. If both strings seem low, then that suggests an inverter fault ..... I think?

I think gefnew's advice is probably the best, contact SMA and see what they suggest. I'd expect them to note the importance of the halving too, and see if there's anything they can confirm/rule out with a chat supervising you while you 'play' with the inverter.

Mart.

PotentialEnergy

Thanks Marty - The problem looking at my SB 3800 is that it is really basic and none of this diagnostics is possible without probably paying for extra add-ons. Easy to wish I'd had them in hindsight.

http:// uk.krannich-solar.com/fileadmin/content/data_sheets/inverter/uk/Data_Sheet_SB33_38_38V-DEN110743W.pdf

I'm going to pose 3 key technical questions for the installer to answer. It'll be interesting what his response is. If he is silent, then thanks to all the help from people, I can quickly get an independent assessment/repair made and then take the matter through the Reccs complaint process I've been referred to by some people outside of this forum.

https://www.recc.org.uk/pdf/consumer-code.pdf?t=201608121628