Inverter Choice: SolarEdge, SMA SB3600TL or SMA SB4000TL?

[JH69]

Hi All,

Apologies for the consumer-oriented questions, but this seems to be one of the best forums on the web for discussing PV.

I’m looking at getting a 4Kwp Solar PV system installed (15 x 265W Yingli Panda Panels), but I have a choice of inverters.

Our house is south facing, approx 45% roof angle, with minimal shading (the neighbour’s roof to the east of our house will shade one or two panels but only in the early morning).

The installer I’m most likely to go with has said he’ll get permission from the DNO to go for the full 4Kw output, and has offered the following inverters:

- SMA SB3600TL
- SMA SB4000TL (£100 premium over SB3600TL)
- SolarEdge system with SE4000 and 15 Power Optimizers (£640 premium over SB3600TL)

I’m aware that the SolarEdge system will probably yield the most Kwh, but given we have minimal shading issues, is it worth spending an extra £640 on? (This is approx 9% more than the SMA system).

If I go with an SMA inverter, which would be the better choice out of the SB3600 or SB4000? I really like the idea of being able output the full 4Kw, but I’ve tinkered with SMA’s Sunny Design software and the projected difference in annual Kwh is only about 30Kwh, so it probably wouldn’t pay to spend the extra £100.

Is Sunny Design a good way to judge potential yield on these systems, or am I likely to see better real-world performance from the SB4000 than Sunny Design suggests?

Is there anything else I should be considering when choosing an inverter?

Many Thanks!

JH69

 

[Vegelen]

Fronius 3.6TL shouldn't be discounted as one of your inverter options.
If you use a transformer less inverter, make sure the inverter manufacturers earthing requirements are met.

 

[jason121]

dont forget STECA, very good inverter like Fronius no bluetooth but only weights 9kg,
If you go with the 4000 must rated to 16amps for G83 or else will require permission from DNO for the 4kw inverters

 

[infinity]

Depends how much shade the neighbours house gives. Its it 2-3 good hrs of sunshine then thats 2-3 hrs of decent output you lose every day! could soon add up, say its 1-2 KWh less a day, X 250 days a year (allowing for dark winter days) soon adds up!

 

[Vegelen]

Depends how much shade the neighbours house gives. If its 2-3 good hrs of sunshine then that is 2-3 hrs of decent output you lose every day!

Time of shading is also important.
Shading in the early morning or later in the evening is less important than around midday, when the suns output is approaching its peak.

 

[infinity]

Time of shading is also important.
Shading in the early morning or later in the evening is less important than around midday, when the suns output is approaching its peak.

Yes true

 

[sedgy34]

We're in the uk are you you could get a few quotes too from here

Also you dont require permission on 1x4kw solar installation
Its only on multi installs in the same street you need permission from DNO
hope your not taking in any spiel!

 

[pww]

Using a 4000TL would require pre approval and really offer nothing over the 3600TL (which is basically a 16A capped 4000TL). Do you want to know what your system is doing on a per panel basis? Do you want to reduce your potential downtime if there is a fault? Do you want an additional 5/10% production? Do you want to minimise what shading issues you have on the system? Answer yes = go with Solaredge! Answer no and to be honest I would use a Fronius IG 3.6TL.

 

[moggy1968]

unless you want blue tooth monitoring, in which case SMA knocks spots off Fronius.
Have used a lot of fronius and very happy with the outputs, noisier than the SMA though I think so depends a bit where it is to be mounted.

 

[Earthstore]

This is for the OP, I fully understand your queries, however your chosen supplier should be specifying the inverter for you and explaining the reasons why.
There are many you could use, and also some can be set with different options to minimise the shading issues that you have.
So my advise would be to choose a supplier that understands how inverters really work, just one example would be how optitrack works on an SMA, and would you benefit from using this facility on your installation.

Without surveying your house it really is impossible to suggest which inverter would perform the best, although for different reasons many have their favourites.

I hope this helps.

Ps Most, not all but most DNO's require a maximum output of 16A, which is 3.68 KW without having prior permission to connect to the grid, some inverters are supplied this way(G83 approved), and some can be set to this.

 

[wstevew]

+1 to Earthstore's comment above.

You need to get the shading modeled to know the impact on yield.

In my opinion the potential downtime reduction of solaredge could be negated by the fact you have more devices (more to go wrong).

 

[Worcester]

We're in the uk are you you could get a few quotes too from here

Also you dont require permission on 1x4kw solar installation
Its only on multi installs in the same street you need permission from DNO
hope your not taking in any spiel!

@sedgy34 if he's going to install an unrestricted 4000TL he will almost certainly need DNO approval to exceed 16A (3.68kWp) and it may even have to be a G59 application now since the gudileines expired in March.

 

[JH69]

Time of shading is also important.
Shading in the early morning or later in the evening is less important than around midday, when the suns output is approaching its peak.

Shading will be minimal, and only early on during the day when the sun is weak. The installer has said he will connect the strings to take this into consideration, so while I expect a SolarEdge system to slightly outperform SMA, in this case, I don’t imagine it would that significant. I suspect around 90% of the time there will be no hard shading, and certainly none when the sun is shining directly on it.

 

[JH69]

Using a 4000TL would require pre approval and really offer nothing over the 3600TL (which is basically a 16A capped 4000TL). Do you want to know what your system is doing on a per panel basis? Do you want to reduce your potential downtime if there is a fault? Do you want an additional 5/10% production? Do you want to minimise what shading issues you have on the system? Answer yes = go with Solaredge! Answer no and to be honest I would use a Fronius IG 3.6TL.

Thanks for the reply.

There are a few things that put me off the SolarEdge system:

1. Extra expense
2. Lack of “Sunny Beam” type portable monitor
3. Need to have the inverter connected to the internet
4. More components to go wrong.
5. If the inverter fails, it needs to be replaced with another SolarEdge inverter.

The thing that I am curious about though is what the extra yield might be on a relatively unshaded roof. I’ve read some suggestions that it can be around 5% or more, but an article in Photon magazine indicates that when there are no shading issues the advantage is probably only 1 – 2% (i.e. not enough to warrant the 9% price premium).

As for approval to use the SB4000TL, the installer has said he has a good relationship with the local DNO and is happy to get approval for the install. For me, it’s really a question of whether it’s worth paying the extra £100 for what will potentially only marginal extra Kwh, but I do like the idea that the system’s output won’t be capped in the way it would be with the SB3600TL

I haven’t looked at the Fronuis inverters, simply because none of the installers have yet offered it to me. How is the Fronius IG 3.6TL when it comes to remote monitoring?

I really like the look of the SMA Bluetooth and the ability to monitor the inverter using either Sunny Beam or a Laptop.

 

[provr6]

If you go SMA, save your money and don't go for the sunnybeam. Laptop does all and more for free (bar the cost of a wireless dongle if you don't have wifi built in to the laptop).

 

[SibertSolar]

How about an Eltek THEIA 3.8HE-t? Built-in webserver, colour LCD, 10yrs warranty, IP65, high-frequency transformer, 97.3% peak efficiency, very low (7W) start-up, integrated DC switch (Santon), output can be capped at 16A. Would be better with 2 strings of 8 modules - could you add another module to your roof perhaps?

Just another option for you (or anyone) - as Earthstore already said, your installer should be giving you options based on your limitations/specification, and justifying each option accordingly. If you do exceed 16A/phase output then you will seriously need to look at a G59/2 approved solution - which in this case hardly seems necessary. However, if you do wish to avail of that little bit extra over and above the 16A/3.68kW (at 230vac) then you could use an Eltek single-phase TL (transformerless) inverter, which are also tested to G59/2:

Eltek 4300TL, 1ph TL, 4.8kWp PV max, IP66, 97.3% peak eff, 7W start-up, embedded data-logger (30yrs), G83/G59 approved, 1 string of 15mods.

Andy

 

[matt25]

If you go SMA, save your money and don't go for the sunnybeam. Laptop does all and more for free (bar the cost of a wireless dongle if you don't have wifi built in to the laptop).

Wait for the "Sunny View" lol

 

[Eco Deal]

solar edge will yeild more per year !!!!

so the extra 600 will return around 5 k extra in 25 years.

so its up to you spend more earn more

we do yingi 265w and solar edge all the time.

 

[JH69]

solar edge will yeild more per year !!!!

so the extra 600 will return around 5 k extra in 25 years.

so its up to you spend more earn more

we do yingi 265w and solar edge all the time.

Hi MEP!

Is that *really* the case if you have no shading issues though?

I had a good read thought the Photon Labs article (http://www.solaredge.com/files/pdfs/se-case-study-photon.pdf), and it seems to show that whilst SolarEdge does produce more even when unshaded, it’s basically only a few percent.

If it was going yield an extra £5k over 25 years then it surely it’s going to have to out-perform other systems by around 25% (very feasible if shading is a big issue, but in my case it’s not).

 

[babba]

Hi MEP!

Is that *really* the case if you have no shading issues though?

I had a good read thought the Photon Labs article (http://www.solaredge.com/files/pdfs/se-case-study-photon.pdf), and it seems to show that whilst SolarEdge does produce more even when unshaded, it’s basically only a few percent.

If it was going yield an extra £5k over 25 years then it surely it’s going to have to out-perform other systems by around 25% (very feasible if shading is a big issue, but in my case it’s not).

Sounds like you answered your own question.

 

[JH69]

This is for the OP, I fully understand your queries, however your chosen supplier should be specifying the inverter for you and explaining the reasons why.
There are many you could use, and also some can be set with different options to minimise the shading issues that you have.
So my advise would be to choose a supplier that understands how inverters really work, just one example would be how optitrack works on an SMA, and would you benefit from using this facility on your installation.

Without surveying your house it really is impossible to suggest which inverter would perform the best, although for different reasons many have their favourites.

I hope this helps.

Ps Most, not all but most DNO's require a maximum output of 16A, which is 3.68 KW without having prior permission to connect to the grid, some inverters are supplied this way(G83 approved), and some can be set to this.

Thanks Earthstore,

To be fair to the installer, I approached him to enquire about pricing for a SolarEdge system, and to ask how it would compare to SMA (as I’d had other quotes for SMA systems).

He’s had a look at the house, and listened to my thoughts and proposed an SMA system with 10 panels in one string, and 5 panels in another. The 5 panel string will cover the area that will get some shading to minimise the effect of it.

He’s said DNO approval won’t be a problem, but I asked him to quote for both the 4000 and 3600 so I could compare the difference in price.

I’m leaning towards the SMA system, but I’m undecided on which out of the two inverters to go for. This thread was about ensuring my thoughts were on the right track.

 

[Earthstore]

The only thing about putting 5 panels on one tracker is it may be doubtful if it would have enough voltage to kick a 4000TL inverter in IE 5x30V=150V, Of course it depends on the orientation of both arrays.
So it depends on the panel voltage really, and total system design.

 

[JH69]

The only thing about putting 5 panels on one tracker is it may be doubtful if it would have enough voltage to kick a 4000TL inverter in IE 5x30V=150V, Of course it depends on the orientation of both arrays.
So it depends on the panel voltage really, and total system design.

Both arrays would have the same orientation. Basically all south facing on the same roof.

 

[Somecamel]

I Thought SMA inverters couldn't handle uneven strings, I tend to use the Powerone 3.6 which is a lot more versatile

 

[SibertSolar]

Uneven strings are fine, so long as they are on different/independant MPP trackers.

 

[Somecamel]

Uneven strings are fine, so long as they are on different/independant MPP trackers.

I stand corrected then Is this just on the 4000TL I assume or does the 3600 do it as well?

 

[JH69]

Uneven strings are fine, so long as they are on different/independant MPP trackers.

Sunny Design seems to default to uneven strings (10 and 5). Anyone know why this is? What are the considerations when deciding how to spilt panels over 2 strings?

 

[SolarCity]

I can't be sure but I would say that this is due to the fact that one of the strings will be running at a higher voltage and will therefore start up earlier - although I would say that the actual benefit would be minimal.

Far better would be to consider the shading (however small) and adjust to suit.

 

[wstevew]

Yingli panda 265w are 31v

4000TL default startup voltage is 150w (can be dropped to 125)

Mppt input range 175 -500v per string.

Voltage out of tracker range?

 

[sedgy34]

Power one for me 2 individual strings on trackers
string of 10 no problem the 5 may be under voltage though would need to run through the design

 

[moggy1968]

Yingli panda 265w are 31v

4000TL default startup voltage is 150w (can be dropped to 125)

Mppt input range 175 -500v per string.

Voltage out of tracker range?

bear in mind output will drop with age, climatic conditions and installation losses.
You don't want to find that in 5 years time your voltage isn't enough to kick start the inverter 'cos you cut your calcs too fine. You need to give yourself a bit of margin

 

[JH69]

Yingli panda 265w are 31v

4000TL default startup voltage is 150w (can be dropped to 125)

Mppt input range 175 -500v per string.

Voltage out of tracker range?

I spoke to both my installer and SMA today about this. They both said pretty much the same thing: If Sunny Design accepts it, then the installation will be ok. While Sunny Design can't advise on what's the 'right' balance of panels on each string - that's location specific and up to the installer to figure out - it will flag it as an issue if there aren't enough panels on a string.

 

[cjapeterborough]

Based on your original post I would cast my vote for the 4000TL as you shouldnt imho design your system to limit your output to the max 3.6KW.

I purchased a 4KW system last June and the various software packages predicted an annual output of around 3000KWh/pa

I have a more or less East facing roof in Cambs and my total for the year was 3600KWh -on cool sunny days peak output hits 4KW

I am sure I have gained an extra 5% at least by not limiting the output to 3.6KW

Depending on the panel spec tolerances you could in theory end up with say a 4.20KW system (some panels are up to +5% output spec)
Dont know the yingli spec but again another reason not to cap the potential of your system.

If you are south of me in the uk I would expect a well designed south facing system to outperform mine by perhaps 10-20%
So the 3600 would hold you back certainly more than the £100 difference or 1-2% inverter efficieny differences
I have a 19 panel Schuco system split 10+9 on each mppt with no shading issues.
Sunny Explorer is all you need for monitoring and there is also additional software for android and such -the sunny beam again imho is
not really needed.

Hope this helps

 

[JH69]

Based on your original post I would cast my vote for the 4000TL as you shouldnt imho design your system to limit your output to the max 3.6KW.

I purchased a 4KW system last June and the various software packages predicted an annual output of around 3000KWh/pa

I have a more or less East facing roof in Cambs and my total for the year was 3600KWh -on cool sunny days peak output hits 4KW

I am sure I have gained an extra 5% at least by not limiting the output to 3.6KW

Depending on the panel spec tolerances you could in theory end up with say a 4.20KW system (some panels are up to +5% output spec)
Dont know the yingli spec but again another reason not to cap the potential of your system.

If you are south of me in the uk I would expect a well designed south facing system to outperform mine by perhaps 10-20%
So the 3600 would hold you back certainly more than the £100 difference or 1-2% inverter efficieny differences
I have a 19 panel Schuco system split 10+9 on each mppt with no shading issues.
Sunny Explorer is all you need for monitoring and there is also additional software for android and such -the sunny beam again imho is
not really needed.

Hope this helps

Thanks!... that's really useful info and it has just confirmed my thougths on the subject: I'm going for the SMA SB4000TL.

I know it will be a rare occasion, but it will be nice watching it peak at around 4Kw.

Interestingly I spoke to SMA today (to ask about # of panels per string), and also asked them about 4000 vs 3600. They said that apart from the capped output they are otherwise the same. So there should be no efficiency differences at lower outputs. The only reasons to not go for a 4000 is the extra cost and the need for DNO approval.

 

[cjapeterborough]

I think they are shipping the 4000TL-21 now which may be slightly improved.
Make sure you get it installed with the latest 3.20 firmware as there are a few annoying
reporting bugs in sunny explorer (inverter doesnt know that sometimes there are more or less than 30 days in a month

 

[Durky]

Solar Edge has the added benefit of a 12 year warranty over SMA at only 5 years. My customers really like the added peace of mind.

 

[Mr.Shing-A-Ling]

The SolarEdge has a 12 year warranty which is another consideration. I have just had a SolarEdge SE4000 fitted with 12 x SunPower E20 333

 

[danesol]

If you go SMA, save your money and don't go for the sunnybeam. Laptop does all and more for free (bar the cost of a wireless dongle if you don't have wifi built in to the laptop).

^^^^ agreed

 

[Eco Deal]

Solar edge as online monitoring which can also pin point under performing panels.

 

[Mr.Shing-A-Ling]

Solar edge as online monitoring which can also pin point under performing panels.

The SolarEdge Monitoring Portal is a nice feature.

 

[danesol]

The SolarEdge has a 12 year warranty which is another consideration. I have just had a SolarEdge SE4000 fitted with 12 x SunPower E20 333

SMA also have an excellent 'out of warranty' 300euro Inverter exchange scheme from their Milton Keynes factory, that should always be considered

 

[finewine]

Based on your original post I would cast my vote for the 4000TL as you shouldnt imho design your system to limit your output to the max 3.6KW.

I purchased a 4KW system last June and the various software packages predicted an annual output of around 3000KWh/pa

I have a more or less East facing roof in Cambs and my total for the year was 3600KWh -on cool sunny days peak output hits 4KW

I am sure I have gained an extra 5% at least by not limiting the output to 3.6KW

Depending on the panel spec tolerances you could in theory end up with say a 4.20KW system (some panels are up to +5% output spec)
Dont know the yingli spec but again another reason not to cap the potential of your system.

If you are south of me in the uk I would expect a well designed south facing system to outperform mine by perhaps 10-20%
So the 3600 would hold you back certainly more than the £100 difference or 1-2% inverter efficieny differences
I have a 19 panel Schuco system split 10+9 on each mppt with no shading issues.
Sunny Explorer is all you need for monitoring and there is also additional software for android and such -the sunny beam again imho is
not really needed.

Hope this helps

90% of the time that my SMA 3600tl is running with optimum sun angle etc I see a figure between 3840w and 3960w and occasionally just over 4000w. It doesn't sit at 3680w even though its capped to this. Perhaps our local voltage is lower and the inverter caps at 16A rather than wattage. When it is clear all day the panels generally get hot and the panels can only output 3300 w - 3500 w anyway so I would hazard that the performance difference is tiny. That said, if we lived a climate that had plenty of sunny cool days such as the winters in the south west states then I think a measurable difference could be shown. Unfortunately cool days in the UK are generally cloudy or the sun is too week during winter months to approach peak output anyway.