Fellow Fishbowlers,

PV Array current is proportional to insolation, with a very weak relationship to temperature. The result is that while DC power is significantly derated with temperature (falling voltage), we should be able to compare Array current at MPPT to the PV manufacturer's MPPT current rating -- if the following conditions exist.
1. The noon time sky is clear (this includeds no cloud edge reflections).
2. The installation angle of the PV array is say within about 15 degrees on normal for the sun.
3. The XR Beta isn't running at its AC power rating, that is running VDC more a few tenths of a volt above MPPT.
4. One is not reading a brief spike of current from capacitors in the inverter.

So, what are my results?
My 48 Siemens SP75s have a rated DC current at MPPT of 53 amperes.

The largest current measured under conditions 1 thru 4 is only 46.5 amperes.

I'm interested in how other PV array performances compare to their ratings.

Gordon

Gordon,

You need a calibrated solar cell and a pyrometer to measure the actual watt/meter/square irradiance reaching the cells , typically , this is in the 900-950wmsq range with the panels perfectly parallel to the sun , the only time I have seen higher is with edge of cloud effect. You are wrong saying 15 degrees will have no effect , the loss is i cos if the angle error , which is not zero.

Also , your panels will degrade with age , If you believe the 25 year warranty , that would be a little less than 1% per year.

You would also need lab quality current sensing , the GP can be off by 2% (component tolerances of the DC sense circuit)

The best test is short circuit testing the panels using a high precision shut and dvm , compare to the short circuit rating and then scale to the pyrometer reading.

Off hand , I'd guess your panels are a few years old , Fishbowl is off little and your NOT at 1000/wmsq and when you put strings in series , each strings current will be limited by the current of the weakest panel. I don’t see anything alarming here , your numbers are only off 12.3% , which the sun angle error is probably about 3% , the irradiance is probably peaking at 950 , so that’s another 5% , which leaves 4% for the panels matching and inverter sensing … and did you get up on the roof and scrub every little spec of grime off every panel ?? … more losses … anyways , seems inline with real world performance to me ..

Henry

Henry,

A couple of thoughts:

1. I current matched the panels in group of four before installation (I was thinking that when you put current sources in series, you get the lowest current of the string). By the way, it is fair to point out that one of the advantages of ST inverter over the Sunboy is the ability to do panel matching -- not practical when the system only has one string as is the case with the SB.
2. I don't mean to suggest that 15 degrees off angle has no impact, my calculation is a loss of 3.4%.
3. On the insolation, keep in mind that I am looking over many days for best sunlight - somewhere, sometime, 1000 watts per square meter and bit more should to be there in my climate.
4. As for cleaning, ok, ok, I'll climb up on my roof, rather than spraying the array with a hose.

Most Important: I testifies to your great work, that I am know able to look this hard at the PVs.

As to others, I am interested in your comparison numbers.

Gordon

Hi Gordon,

The 1000wmsq is fiction , I have never seen this without clouds .. ever. I also looked at dataloged sites with calabrated equipement , again it almost never happens .. best peaks on clear days I have ever seen is about 960wmsq

Smart matching your panels and YES , I aggree this will further narrow the difference on a SB to ST the long string is much more likely to have mismatch issues

One more thought ..

In your climate ... you always have some amount of smog , yes ? ... this will reduce the wmsq ...

Henry,

Some smog, not anything like LA. I'm in San Diego County about 23 inland.

I find it interesting that insolations above 960 are hard to come by in clear.

The matching issue is quite significant, if you order 24 modules (in my case 48), one can very nicely match sets of four. With the SB, and 24, all you can do connect
all in series and produce the current due the weakest module. In my group of modules the difference was about 3.5 %. I also suspect that the long string used afor the SB sets up a long-term problem because which ever module degrades the fastest, takes the current for the whole PV system with it.

Gordon

alot of factors effect isonalation

Time of year , the lower sun angles , more atmosphere it travels thru.

The actual elevation , you and I are near sea level , as is most of Florida , which is the data I have looked at. I'm sure there is government data some where online that has this data , specific to your location , my point is 1000wmsq is not a daily clear day thing

I don't know why the module thing never clicked on in my mind before ... but yes , it would most definantly effect Sunnyboy installations ... Also recently the comparision what happens with partial shading on a multistring Sunnboy is much worse than same thing on ST.

This is a real concern , now that almost all panels greater than 120 watts are 24 volts , this means most Sunnyboy 2500 installations will need two strings

Henry,

Consider that the array current produced by 24 modules in series isn't just set weakest module, worse it is set by the weakest cell. So, a if bird leaves a "bird spot" the size on a quarter on one cell out of the 864 cells in the string used for a SB inverter (of course it makes no difference if the setup uses nom. 12 volt or nom 24 volt modules, the cell total is 864) I would expect a reduction in current for the entire propotional to loss of light to the one cell!

Take the same 24 modules and the same "bird spot," and wire them into 6 circuits of 4 modules each for an ST inverter, I would expect only 1/6 of the array current loss.

There is more to the ST vs SB comparison than simply efficiency numbers, you and Xantrex have the superior product for the real world. I just hope that Xantrex Marketing can sell that point to the world.

Gordon

The single or dual strings on an SMA inverter is not as bad as it seems at first. Each module or even strings within a module have bypass diodes installed to protect the cells from reverse bias, as may happen when you get a shaded cell. This will allow power to pass by the shaded cells, reducing the voltage but letting the amperage through. The Sunny Boy inverter has a wide range of voltage acceptance, so even if one whole module is shaded it will only reduce power by that module, or there abouts. I've seen an array of Schott 165's where we shaded up to 2 or 3 modules with only the proportional loss of power. While it wasn't very hot when we did this, and it would have more effect with this array in hotter conditions, it will depend on the array makeup how many modules could be shaded before you lose the array completely. If you had 24 SP75's in series, you could shade quite a few before losing it all. On top of that, for each shaded cell you only lose the portion of the module bypassed by that one diode. If you have a 48 volt system, losing that one string of cells will almost certainly drop the entire 4 module string.

I'd like to say that I'm very glad to see the ST inverter has been reworked into what appears to be a top notch unit now. I hope that it proves reliable and offers one more good choice in PV equipment to the consumer and installer, thanks to Henry [img]images/smiles/icon_smile.gif[/img]

I aggree a single string sunnyboy is really good at handling partial shading.

What Gordon and I are talking about is module tolerance ... if a single module is say 3% low on CURRENT , then the whole string suffers , not so on the Suntie

Thanks for the positive comments on my Suntie work [img]images/smiles/icon_biggrin.gif[/img] [img]images/smiles/icon_biggrin.gif[/img]