Thoughts and feedback on connecting panels together: which on what string?

Fenachrone

Right now my array consists of five Solar World modules:

Specification SW 235 Poly
Peak Power Watts (PMAX) 235 W
Maximum Power Voltage (VMPP) 30.0 V
Maximum Power Point Current (IMPP) 7.85 A
Open Circuit Voltage (VOC) 37.0 V
Short Circuit Current (Isc) 8.35 A

...and one GMA Solar module

Specification GMA GMAP6-60-270W
Peak Power Watts (PMAX) 270 W
Maximum Power Voltage (VMPP) 30.94 V
Maximum Power Point Current (IMPP) 8.80 A
Open Circuit Voltage (VOC) 39.26 V
Short Circuit Current Isc) 9.41

The GMA was a replacement after destruction of one of the Solar Worlds with hurricane Fiona. It works fine in the current array setup,
which is two strings of three Solar Worlds and then two Solar Worlds and the GMA. 

This current array is inadequate for proper charging, especially in early spring and late fall.  Consequently, I will be adding two Trina Vertex S 395W panels.   

Here's the Trina specs:

Specification: TSM-DE09C.07-395
Peak Power Watts (PMAX) 395W
Maximum Power Voltage (VMPP) 34.0V
Maximum Power Current (IMPP) 11.62 A
Open Circuit Voltage (VOC) 41.0 V
Short Circuit Current (ISC) 12.21 A

I'm contemplating one possible scenario for hooking these new panels up:

1) the two Trinas together in on one string

2) the five Solar Worlds and the GMA on the other string.

My rationale (though not the product of any serious technical know-how on my part) is that the

electrical characteristics of the Trina panels won't mix with the rest. 

Any feedback on this is greatly welcomed.

BB.

You do have the option of adding a second MPPT charge controller--One for the 5+1 array, and the second for the new 2x Trinas.

The rough rule of thumb is that adding panels in series should have Imp between series connected panels be within 10% Imp (i.e. 10 amps +/- 1.0 Amp).

And for panels connected in parallel Vmp-strings should match within 10% (i.e., 60 volts +/- 6 volts).

The Trina panels are Imp=11.62 amps. The Solar World Imp=7.85 amps. The GMA is 8.80 amps...

The GMA and Solar World are not a great match--But are probably close enough to work OK (you one GMA is not operating at optimum).

The Trina's are larger Imp than would be recommended for either SW or GMA series connections (11.62 vs 8.80 vs 7.85 amps Imp).

The Vmps are SW 30.0 volts GMA 30.94 and T 34.0 volts... The Trina Vmp is a bit high to "match" with GMA... But you have what you have, so say they are close enough match for Solar.

Basically, you have 5+1+2 panels or 7 total... What is the max Vpanel input for your charge controller. For example say 8 panels * 30 volt Vmp = ~240 Vmp-array--Most "typical" solar controllers have around 140-150 Vpanel-max specs (at standard temperatures). If you are around Greenfield Mass, your lows get down into the 10's -- Could enough that we have to worry about Voc-array being higher in cold weather.

Anyway--Lets say that the max Vmp-array would be around 110 Volts (just to keep things "simple" for now--Voc-array less than 140 volts at 0F).

110 Vmp-array max / 34 volts Vmp Trina = 3.2 panels = 3 panels in series max...

That would be:

3 SW  in series (Vmp-array = 3 * 30 = 90 volts)
2 SW + 1 GMA in series (Vmp-array = 2 * 30 + 1 * 30.94 volts = 60.94 volts)
3 Trina in series (3 * 34 = 102 Volts Vmp-array)

If you used 2*34 volt Trina, that gives you 72 volts for that one string--It would not work with the other 3xpanel strings.

So, the first suggestion would be to get 1x more Trina solar panel.

The other questions are Brand/model of MPPT solar charge controller, and battery bank Voltage... Want to make sure you X panel array works well with your charge controller + battery bank... You can over panel MPPT controllers, and they will work fine/reliably. But they could "clip" the output charging current to the battery bank during sunny weather (clip to controller's maximum output current rating). Which would be a bit of a waste (add more panels, but controller "clips" (or limits Array Wattage) output current to battery bank during sunny weather).

Configuring an array for MPPT controllers is a bit of a pain/confusion. There can be multiple solutions of series/parallel, or "no solution" in this case with just 2+ new Trinas) that make sense.

-Bill

Fenachrone

Thanks for your very detailed comment, Bill.   I'll add some information to answer your questions:  The MX60 charge controller will handle 150v input. 

The GMA and the Solars together seem to be working OK, but yes, probably some compromise there but appears negligible.

I see the point in adding in another charge controller.  What I want to do is get through this fall then install a higher capacity cc.  

As far as weather conditions overloading the array, this is a seasonal residence and is shut down by the beginning of November -- everything goes off.  Charging conditions in Atlantic Canada at this time are pretty weak sun and lots of overcast. In summer I have the option of shutting off one string, but this year we had few consisent, clear, bright sun days, and this seems to be more of a weather trend  up there.

I have the room for one more breaker in my combiner box so could add another string. 

BB.

Midnite Solar has a nice array calculator... You can plug in your array information and min/max temperatures (Voc-array-cold and Vmp-array-hot) to figure out if your "combination" will work...

For example, very roughly, Voc-std=41 volts will hit around 141.3 Voc-array-cold (-20C). It appears that you may be able to run 3*Trina panels in series and not exceed 150 Vpanel-max input rating for MX60.

You do have another voltage "issue":

https://www.solarelectricsupply.com/media/sparsh/product_attachment/custom/upload/MX60_a.pdf

PV Open Circuit Voltage (VOC) 150 VDC absolute maximum coldest conditions / 140 VDC start-up and operating maximum

So, it is "on the edge" where a "frozen array" may not start until more sun hits the array and "warm it up) a bit (which is "normal" in the sense that solar panels always heat up when under "significant" sunlight).

More or less, a good "optimum" over paneling of an MPPT solar charge controller is around 77% (derating for hot weather panels + controller losses). For a 60 amp controller on a 48 volt battery bank:

• 60 amps controller maximum * 58 volts batt charging * 1/0.77 over paneling = 4,520 Watt array

This will cause some "clipping" of array output (limiting output to 60 Amps) on cool/clear days for a few minutes or an hour pretty rarely (about the maximum output most system will every experience).

Your "composite array" of 2,595 would work fine on a 48 volt battery bank--If 24 volt bank:

• 60 amps * 29 volts charging * 1/0.77 panel+controller deratings = 2,260 Watt array "optimum" @ 24 battery bank

If you have lots of cloudy weather, a 2,600 Watt array would not be "that much over paneling"--So could make economic sense. However, there is always the issue of "miss-match" between:

• SW 3 series * 30 Vmp = 90 volts Vmp-array SW
• Trina 3 series * 34 Vmp = 102 Volts Vmp-array Trina

You could always try the combination and see how it works. Do you get somewhere close to 77% maximum array wattage (probably over 67% is probably "really good" for the miss-matched strings). It won't break anything.

The issue is that MPPT controllers are looking for peak array power (P=V*I)... You don't know if it will "pick" 90 volts, or 102 volts or something in between (there are typically at least 3 "local peaks" and typically one "absolute" peak--Many--especially older--controllers may miss the absolute peak). When you have 3 different strings with different Vmp-string voltages--It is "messy" from a software engineering viewpoint.

The "power peaks" are rounded--So a bit of peak data smearing may work in your favor (keeping "miss-match losses" to a minimum).

Here is a more detailed discussion on mixing/matching solar panels:

https://forum.solar-electric.com/discussion/5458/two-strings-in-parallel-with-unequal-string-voltages

-Bill

Fenachrone

Thanks again for all this input.  My battery bank is 24 volts, BTW.   So, to recap, per your comments, the best option might be:

3 Solar Worlds in series
2 Solar Worlds + 1 GMA in series
3 Trinas in series

I'll be upgrading the charge controller next spring to a FLEXmax 80.

BB.

That is certainly where I would start...

Monitor the system in full sun and a discharged battery bank (80% or less State of Charge), And see that you are getting (at least) 50%-77% of the solar array's name plate wattage (cool, clear day, sun "square to array").

If you do not get "expected wattage", you might be a) harvesting mostly from 5+1 SWGMA strings, or from 3 Trina, or 3T+2+1SW, etc...

You can revisit the size of your array vs battery bank (suggest 10%-20% Rate of charge for "typical" off grid system--I.e., 200 AH * 10% = 20 Amps minimum)... Or for a 60 Amp charger+array = 300-600 AH @ 24 volt battery bank...

If the battery bank approaches or exceed 800 AH, you might wish to look at a 48 Volt battery bank (solar charge controller will work, however would need new AC inverter, and other DC equipment).

Long term, you may wish to keep the MX60 for the 5+1 array, and a "new controller" for the next array. Whether a full size MX80, or just "large enough" for your 3x Trinas--Your choice.

Be aware, this mix/matching of solar panels is a constantly moving target. MFGer's keep changing wattage/Vmp/Imp for panel designs... If you want a larger array, it may make sense to get the set of Trina's now--So you don't have "an issue" with a mix and match between the MX60 and MX80 arrays in the future.

-Bill

Fenachrone

It also just occurred to me to perhaps follow the K.I.S.S acronym, and just add two more of the readily available GMA panels.  That would compromise a couple hundred watts, but not a big deal ultimately. 

BB.

Yep--It is just a choice for mix and match of what panels are available at a cost effective price.

-Bill