How to properly build a solar string for a grid tie

[imsmooth]

It's been a while since I've been here. I currently have 12.25kw of solar panels on my roof that were installed professionaly. I've build a grid-tie wind turbine that I did all by myself.

I am thinking about adding another 1kw of solar panels on my property, but I would like to do it myself. I would like to know the correct way to size the panels for a Sunnyboy grid tie. This would account for the proper input voltage and current to the unit.

Here is the sunnyboy tech page. I am using the 700W unit as an example.
http://www.sma-america.com/en_US/products/grid-tied-inverters/sunny-boy/sunny-boy-700-us.html

I am looking at a web page of solar panels with the specs here
http://www.sunelec.com/

This is what I am thinking, so please let me know if I am wrong:

If I set the inverter for 750W it says the voltage range is 100-160v, maximum current is 7A. Vmax is 200V
The Yingli 160W panels have a Vmpp of 23V @7A. So 6 tied together is series gives 138V, which is in the middle of the range. Since this is series the current is 7A. Both of these numbers are correct for the inverter and setting.

Is this correct? Is this optimal? Does the Vmpp and the Voc in the series string have to both be within range of 100-160V, or does the Voc just have to stay below the inverter's maximum voltage, which is 200V in this case?

[DamonHD]

Voc, especially when the panels are cold and Voc is higher, must always remain below the inverter's Vmax or the magic smoke will escape, expensively.

[Note for other readers: most places require someone with the right certificates to attach microgeneration equipment to the grid, and that equipment must often by type-approved.  This is in part for safety reasons.]

Rgds

Damon

[imsmooth]

Ok, Voc must stay below the inverters Vmax. Should I strive to have the string's total series voltage be close to the middle of the MPP range, or should it get as close to the max without going over? I understand about leaving a margin of error so I don't blow the inverter.  I am guessing that the higher the series voltage, the sooner it will turn on when the sun is minimal.

[DamonHD]

It can be slightly tricky depending on the efficiency curve of the inverter, the power curve of the panels, their angle and orientation, any shading and your typical weather.

The PV pros have programs to simulate all that stuff, including the local weather, and pick the best string length and indeed the best inverter to maximise yield.

You might look at the efficiency curve of your inverter and try and chose Vmp to be close to the 'best' voltage at typical daily temperatures, being careful about Vmax.  That should get you to within a few percent of best-possible, at a WAG.   Note that I am not a pro, and I'm not allowed to sell you financial products either!  B^>

Rgds

Damon

[Rob Beckers]

imsmooth, generally the strategy is to look how many panels fit in a string while staying below the maximum voltage of the inverter. With that done, the MPPT voltage of that string under normal conditions is much lower, and more often than not is within the range for the inverter. Keep in mind that Vmp goes down as panels get warmer, and the number in the spec sheet is at 25C. Panels outside in the sun are 20C above ambient (at the least), so generally they are far warmer than 25C and consequently Vmp is quite a bit lower.

A little calculation showing how it's done may be helpful (for anyone doing string sizing): For your Yingli 160W modules Voc is 29.0V @ 25C. Their temperature coefficient for Voc is -0.37% per degree C.

If you live in the same climate as me, the absolute minimum temperature measured in winter is -35C. To account for colder-than-usual weather, cold spots, radiative cooling etc., I like to have another 5C margin, so I use -40C for our area (there are various opinions on how cold one needs to go for inverter safety, but let's go with this). So, we have a temperature difference of 25 - (-40) = 65C.

This works out to a voltage change of 65 * 0.37 = 24.05%. In other words, on the coldest day the panel's voltage could go up to 1.24 * 29.0 = 36.0 Volt!

The maximum voltage for the inverter is 200V, so that makes for a maximum string size of 200 / 36.0 = 5.5 panels. They don't make half panels, so your maximum string size is 5 panels if your climate is similar to ours.

With 5 panels in series, Vmp for the string @ 25C works out to 5 * 23.0 = 115 Volt. As you can see, that's barely above the lower cut-off of 100V, and actually so close I would probably not use this inverter in our climate since the string MPPT voltage will drop even lower on a hot summer day.

To show the latter, Vmp moves at about the same rate as Voc, so -0.37% per degree C (not quite the same rate, but close enough for government work). On a hot summer day here it gets to be 35C, and the panels on a roof will be around 60C. That means the voltage changes (60 - 25) * 0.37 = 13% lower, or 0.87 * 115  = 100 Volt. Right at the cut-off for the inverter...

In short, for our climate here the shortest string that can be used is 5 panels (to stay within the MPPT voltage range of the inverter). The longest string is also 5 panels (to stay within the maximum inverter voltage).

Hope this helps!

-Rob-