SB 4000TL - excess power

[claypole]

The Sunny Boy 4000TL spec gives max dc power in 4200w. Can anyone enlighten me as to what happens when your array size is greater than 4.2kwp? Or more specifically when the dc power in actually exceeds 4.2kwp. Sunny Design result for 4.9kwp is more than acceptable, because presumably the array is harvesting more energy for 90% of the time and will only rarely peak above 4.2kwp. But what does the inverter do at these times? Does it simply clip the power (in which case where does this power go?), does it shut down? or does it somehow electronically feather it by rapidly switching on/off (forgive my lack of technical language). And further, would this process result in a shortened life for the inverter?

 

[claypole]

Sorry, don't know why this thread has duplicated and don't know how to delete dupe. Help!

 

[edexlab]

I've just done this on sunny design out of curiosity
sizing the array at just under 4.9kw 26 x Sharp 180s on a 4000tl and the nominal power ratio is 179% although it says conditionally compatible, this is not a good ratio between array and inverter, and if you go back over the design, the actual result is based on 2 inverters with 13 modules on each the 4000tl is far too big for a small array like this
Its always worth checking the design ,rather than taking the result at face value!

 

[claypole]

edexlab - thanks for your input but rest assured I've been playing with SunnyDesign long enough to watch out for changes to number of inverters or panels. If you manually change the inverters to 1 and then check the result for 22 x Sharp 180 (3.96kwp) against 26 x Sharp 180 (4.68kwp) (both configured as 2 strings) you should find the resultant kwh/kwp figure is higher for the 4.68kwp configuration.

 

[claypole]

Well, for those interested and who didn't know already I've found a great thread from down under which explains the answer I'm seeking, albeit with some divergence of opinion.
- Overpowering Inverters - Energy Matters Forum

Nice to see they also have TIC / DNC definition problems down there too!

I can see now why the Sunny Design is showing such good results for undersizing the inverter. In the summer months when the sunshine hours are longer and proportional kwh therefore much greater, the panel outputs will be significantly less due to increased cell temperature. Fairly typically this would be about 10% less. So a 4kwp array might only run at 3.6kw for extended sunny periods, before further loss through the inverter. A 4.7kwp array (which might appear well oversized for a 4000TL) would in fact be running at close to 4.2kw which after inverter loss would reduce to 4kw - the max ac output of the 4000TL. As an additional bonus I presume the increased voltage from having longer strings would also extend the output at beginning and end of every day. You would obviously loose out on a cool bright winter day when the 4kwp array is at rated output and the 4.7 is being severely clipped but for how many hours would these ideal conditions occur over a year compared with the number of summer sunshine hours when the larger array is outputting at 100% inverter output? I would presume the summer voltage drop would be somewhat higher in most parts of Aus, but given the NOCT figures on most panel data sheets I think 10% power reduction for our summers is not far wrong.

 

[edexlab]

can you put up a link for this article ? also on the basis of this are you planning on oversizing an array and did you get any info on the effects of this on the inverters lifespan?. I think it would be worth contacting Sma and see what their opinion is.
As for the 10% reduced size of inverter to array in UK that was the general advice in books /courses (in UK and Germany at Sma) I've done, they based this on designing a cost effective system rather than having yield losses
this shows that the 4000/5000 TL is quite flexible in terms of array variations possible ,we've used a few on 4kw domestics and spec'd them for three phase jobs around the 10 to 15 kw mark ,but never actually got to do this as we used tripowers instead and so not really had any jobs where this setup would have been used

 

[claypole]

Sorry, I thought the 3rd line in my above post was a link but maybe it's just a shortcut on my pc, full address is :
- Overpowering Inverters - Energy Matters Forum
http://forums.energymatters.com.au/solar-wind-gear/topic1617.html
(Actually I think the original link should work because I'm currently on a diff pc and it works fine)
It's a long long thread and I must confess I haven't read it all but most of the salient points were in first few pages. I contacted sma direct as this was my main concern but after a few days I just got a stock reply requesting array design and they would suggest suitable inverter, completely ignoring my question. I would like to get a definitive answer but failing that I might go down the extended warranty route to cover my back. Sunny Design result is conditional compatibility so I don't think sma could void the warranty. I'm planning a 3.92 array but then extending to 4.9 next year. I will go 4.9 now if I can get my installer on board to state TIC and DNC as 4kw on FIT registration - based on the 4000TL max output. Otherwise I want to leave a year gap so the original 3.9 gets highest FIT rate band. I'm guessing with 3 phase the tri power is better option to ensure even spread across each phase which would enable you to max out power under GN83/1.

 

[claypole]

I forgot, in that aussie thread about 3 or 4 pages in someone has posted a graph which plots V and A during clipping and clearly shows the inverter pushing up the voltage which in turn reduces the A and due to the ic curve reduces the power. I don't fullly understand how this works but I guess it's the same way that am MPP tracker functions. The question then is perhaps does this clipping (over voltage) degrade the panels sooner?

 

[Worcester]

@Claypole, Just PM'd you, Gordon

 

[Energetic]

Got a message on a 4000TL today stating: MOD GRID PARAM NOT POS during commissioning. It will not accept the G83 settings (5,8). All the manuals came in french... Have i got a french inverter that will not speak to our grid? it doesn't work on any other settings either as far as i can tell... setting 0,1 and 0,0 on the dials states: 'parameters not changed' possibly with 'grid' at the front, cant remember now... Anyone had this stuff from a Sunny before? They've all worked first time before...

 

[t4sltd]

If DC power is greater than rated input the inverter alters voltage to keep current within limits, thereby limiting power. This is why Sunnyboy design will let you put 4.9kW on when is says 4.2kW max in; the system will not convert all energy in the height of summer when it backs off but work very well overall. 10-20% oversizing of panels to inverter used to be the rule of thumb for system design.

 

[claypole]

Energetic - see emglish install manual here http://download.sma.de/smaprosa/dateien/5692/SB30TL_40TL_50TL-IEN103630.pdf from excellent sma site here SUNNY BOY 3000TL / 4000TL / 5000TL. SMA Solar Technology AG
It's my understanding that uk units should be preset to G83/1 and the dials won't alter that set parameter without Grid Guard protection code. Are you sure 5 8 is correct? It's G83/1 but it shows grid guard not set and I'm sure the std uk parameter is for grid guard set. Do you know what dials were on originally from box?

 

[Energetic]

They come set to 0,0 normally but there is a seperate leaflet in the box stating the UK G83 settings (5,8) that should be applied. Not all the units supplied in the UK are UK units. This one is a french one going by the manual but the g83 settings can be specified on the dials.

Have had a look through the manual and spoken to the tech guy at our supplier and he's not heard of this happening before. We'll be chasing it up with SMA tomorrow.

 

[edexlab]

Got a message on a 4000TL today stating: MOD GRID PARAM NOT POS during commissioning. It will not accept the G83 settings (5,8). All the manuals came in french... Have i got a french inverter that will not speak to our grid? it doesn't work on any other settings either as far as i can tell... setting 0,1 and 0,0 on the dials states: 'parameters not changed' possibly with 'grid' at the front, cant remember now... Anyone had this stuff from a Sunny before? They've all worked first time before...

I've had this on two 4000 TLS( both yellow in colour although I'm not sure thats any indicator of they which country they are meant for) these both had German manuals only
Turn it off wait a while set the dials like you said A 5 B 8 recomission , thats worked both times although its worth setting them on zero then to 5/8 as the rotary switches can look correct but still be slightly off .
I had to speak to Sma on tuesday and found them to be very helpful

 

[claypole]

​

​

"The results of the G83/1 tests are summarized in this certificate. SMA declares that all devices (with G83 setting) that are shipped to the UK comply with the requirements defined in engineering recommendation G83/1. These setting cannot be changed by an installer, user or by any other person without the use of a tool (password protected). Complete documentation on test details are available at SMA on demand."

The above quote is taken from the G83/1 certificate for 4000TL http://download.sma.de/smaprosa/dateien/1807/G83-1_4000TL_20-ZEN092811.pdf
If the inverter you have is not protected from re programming parameters is it perhaps the case that it is no longer G83/1 compliant?

 

[edexlab]

no I spoke to Sma they checked serial no and told me what to do and that as long as its set correctly on dials that it would conform to G83
but I suppose it may be possible that someone could set it to another countries settings which may have different parameters which would make it non compliant but thats down to the Installer to check

 

[claypole]

Can't find the source now but I thought under G83/1 if the inverter was programmable it had to be made tamper proof, as per the sma certificate which certifies that all UK stock is password protected to prevent alteration. So if the dials are in fact re-setting the country paramaters without the need of password then it is no longer tamper proof

 

[Energetic]

I know what you're saying.The G83 commissioning declaration for our DNO states this:

"The protection settings are protected from alteration except
by prior written agreement between *DNO* and the Customer
or his agent."


Which they arent for G83 UK settings... Unless you use superglue... Why would anyone want to alter them tho? Oh right, i think i know...

 

[claypole]

But see again the wording on SMA's G83/1 certificate for the 4000TL
http://download.sma.de/smaprosa/dateien/1807/G83-1_4000TL_20-ZEN092811.pdf
I think any unit intended for UK distribution should be preset at the factory and then Grid Guard set so that you can twirl the knobs any way you like but they won't alter the G83/1 settings. If moving your knobs is changing the parameters (without password use) I think it indicates that the unit was not preset at the factory for UK use, which is perhaps supported by the lack of an english manual in the box. I think the 5 8 setting will set it to G83/1 but not with Grid Guard set, ie not tamper proof as a factory preset UK unit would be, which means strictly it would not satisfy your DNO's request.

 

[confounded]

Energetic - see emglish install manual here http://download.sma.de/smaprosa/dateien/5692/SB30TL_40TL_50TL-IEN103630.pdf from excellent sma site here SUNNY BOY 3000TL / 4000TL / 5000TL. SMA Solar Technology AG
It's my understanding that uk units should be preset to G83/1 and the dials won't alter that set parameter without Grid Guard protection code. Are you sure 5 8 is correct? It's G83/1 but it shows grid guard not set and I'm sure the std uk parameter is for grid guard set. Do you know what dials were on originally from box?

We have had a similar issue to Energetic and having spoken to SMA they are issuing the grid guard code so we can set the inverter from default to the UK setting via Sunny Explorer. Anyone have to do this before?