East - West Split - How many invertors???

[Worcester]

Thought you'ld like thise:

The Front page of the PV*Sol reports:
1 x SB3000TL-20 Output to Grid = 2042.8 kWh/a
2 x SB1700 Output to Grid = 1987.0 kWh/a

It'll proabably be down to inverter efficiencies. In theory East/West, equal strings is no different from two strings same orientation, as a shadow passes over both strings, however with only one MPP tracker ... it's not what I expected!

 

[SolarCity]

Goes against what I've been told also. I have correspondence from SMA where they strongly suggest dual MPP tracker inverters even in cases of shading. With east-west arrays, I am very surprised that the result comes out better for the single tracker inverter over the two seperate inverters.

In fact, aside from the inherent better efficiency from TL inverters, I can't work out why the results come out like they do.

 

[claypole]

Intertesting. Is this partly due to the global peak mppt used in the 3000TL? Made for shade. SMA Solar Technology AG
Is it also fair to say that the E/W split problem is overplayed? If the sun's arc is 180deg and each roof roof face is pitched at 40deg that still leaves 100deg (in the best part of the day) when the sun is directly accessing both faces. And even when it's not directly accessing would the lower cell temp on the shaded side make up at least some of the ground lost by lower insolation? I'm also surprised by the figures to the extent of questioning if they can be right but now I'm nt so sure.

 

[claypole]

Gordon, can you maybe run a 4000TL through pvsol. I realise it will show oversized but I'd expect still to see a higher output with the advantages of twin mppt and the TL higher efficiency. The start voltage is the same as the 3000TL but the mppt range is actually slightly wider. I can't see why it wouldn't give the best output, balanced of course against the extra cost, but then once you've passed the payback period surely you want the highest possible output assuming you can fund the initial spend.

 

[edexlab]

Edexlab, thanks for this comment because I fear I may be at error. The table I had been looking at is here http://download.sma.de/smaprosa/dateien/7418/Duennschicht-UEN093821.pdf which I'm guessing is the same table you had found. It does not of course mention Sanyo Hybrid. At the time I was looking at it I was considering the hybrids but from their data sheet I understood them to have amorphous Si(superstrate design) which would make them unsuitable for the 4000TL. I should have re-checked the table before I posted because I realise I've written the comment incorrectly as if they are mentioned by name. If you also believe this is factually wrong and that they are in fact suitable please let me know and I'll edit the post.
Interesting material here too http://download.sma.de/smaprosa/dateien/7418/Ableitstrom-UEN092510.pdf

HIT Hybrid cells are designed to" prevent the free charge carriers recombining on the back electrode ,
this is the main cause of capacitance discharge"
they are not the same as thin film but do use some of the same materials but have different characteristics

Also Sunny design highlights modules where a -/+ ground is required

 

[claypole]

Thank you edexlab. I can't see how to edit post so will just advise here please ignore erroneous info on post 10 ref 4000TL / Sanyo Hybrid compatibility.

 

[Worcester]

Gordon, can you maybe run a 4000TL through pvsol. I realise it will show oversized but I'd expect still to see a higher output with the advantages of twin mppt and the TL higher efficiency.

PV*SOL throughs a wobbler as it oustside the permissible sizing range - see attached (click on it to enlarge), so it won't run the economic simulations!

 

[SolarCity]

Gordon, did you adjust the sizing parameters on your version of PV Sol? My version only likes 90% - 120% before it starts to panic.

 

[edexlab]

Thank you edexlab. I can't see how to edit post so will just advise here please ignore erroneous info on post 10 ref 4000TL / Sanyo Hybrid compatibility.

No problem we're all here to learn I can't tell you how much this forum has helped or even just caused a bit of doubt which is no bad thing as most people will then go and check for themselves and the correct info is brought to light
I have been on other forums and there is always someone on there spouting rubbish and no one questions it .
but this is the best one I think with some really good participants

 

[Worcester]

@BiggSolar,

I always let my system choose by spec and location, here it is tweaked to allow 80% sizing, output is very similar to the SB3000TL-20, even with the twin MPP.




My guess is that with two equally sized strings, where the primary generator is being slowly switched through the day from east to west with a balance at or around miday when both will be getting similar irradiance, is that the system is just coping.

We need someone to set this up with all three invert configurations and then measure the voltages across the strings and the current through them over a 12 hour cycle and then plot it out along with the actual Wh (not kWh) generated by the inverter at the same time..... (of course in the single tracker set up, I would expect both string voltages to be the same, though with different current flows)

 

[claypole]

Fascinating and I fear I've made a schoolboy error in assuming that an E/W must benefit from 2 inverters or at least 2 MPPT's. If you look at a typical panel power/voltage curve the peak power voltage point is very similar across the full range of irradiation. If you then also consider that the diffrence in irradiation between E and W at any given time will never be huge (because line of sight sun is only 1 part of the overall irradiation) their max volt points will be even closer. And provided we have 2 strings in parallel the different currents will be aggregated. I think the only problem would be with 1 long series string where the sunny side roof output would be pulled down to the same level as the shady side - or am I making another schoolboy error.
I would suggest the 3000TL outperforms 2x1700 perhaps because the start voltage is slightly lower and mainly because of overall efficiency difference, (Max 97.0 v 93.5, and Euro 96.3 v 91.8), bearing in mind that from Gordon's pvsol calcs I think there is less than 3% difference.

 

[SolarCity]

I've just been designing a system with an East/West split and thought I'd let you know my results so far.

Using 16no. 250w Sanyo panels, all on the east side of the roof, using an SB3800 inverter and using the climate data from Plymouth, a yield of 3,258.1 kWh is predicted.

Using 16no. 250w Sanyo panels, 8 on the east side and 8 on the west side, using two SB2000HF inverters and using the climate data from Plymouth, a yield of 3,123.6 kWh is predicted.

I've tried a few different inverer combinations but still get similar results. PVSol Expert consistently returns a better yield on an East or West facing roof rather than using a West/East split.

For comparison, if the same roof was adjusted to face due south, the same array would return 4,042.7 kWh. This means that the east facing array works at around 80% of the efficiency of the south facing array (consistent with SAP figures) and the West/East split configuration performs at around 77% of the optimum. Roof angle with all calculations was 37 degrees.

My conclusion so far is that an East/West split is useful when space is an issue, but not a good option if the goal is purely to maximise yield - not least because an East/West option is more expensive (dual MPP trackers/two inverters, extra scaffold).

 

[Worcester]

@BiggsSolar, Ireckon you're right, in a simple installation, one big East facing array would be best, if not then spilt with either dual MPP's or two inverters.

So here's the big one,

And it goes back to @Claypole's initial discussion of overpowing inverters.

How can I maximise the power I generate from my roof (for the sake of discussion assume cost irrelevant) when it faces East / West and staying within the G83 16A limit for as long as possible during the day?

The two arrays/strings have mirror image power generation profiles (morning / night), with peaks probably being very close to each other around late morning and early afternoon.

So what would be the optimum/maximimum size I should put on each side?
And which inverter(s) perform best when overpowered to clip at the combined output of 16A. I would suggest it would be a single inverter (that way 16A can't be exceeded) with two MPP's, however wether it clips or 'cuts out' is probably the most important issue, and just how big can those arrays be?

 

[SolarCity]

I'd imagine that the big problem with these kind of set ups is the midday period, where both arrays would be getting lots of sunlight. I can't run this through PV Sol but I'd imagine that using a 16a limited inverter with a combined east and west array of much more than 4kWp would experience losses during this peak period as the inverter 'clips'.

In reality, this isn't really an issue because the 4kWp threshold on the FITs prevents us from installing in this way anyway.

 

[Energetic]

Can you run it through on a 4000TL on separate MPPs. Might this help?