# Another weird setup

Options
Solar Expert Posts: 336 ✭✭✭
HI.

We've just visited an existing installation, where the owner has cleverly put 2 grid-tie Kyocera panels similar to the following: http://www.kyocerasolar.com/assets/001/5129.pdf .... connected to a 24V charge controller and 24V battery sytem.

The Vmp is around 26V in STC and only 24V in NOCT, and what I cant understand is that in theory the system shouldnt work, yet it seems to be charging his batteries somehow, maybe not optimally, but charging it is.

Can anyone enlighten me on how this may be happening?

thanks

Options
Re: Another weird setup

It turns out that the peak Power/Current curve is fairly flat... So while Vmp~24 volts, if you look at a generic (Suntech) panel I/V/P curve (PDF), if you are willing to accept some reduction in current, you can get higher voltage from the panel:

Vmp-noct~35 volts at Imp-noct~4.18 amps... If we decrease current by ~1 amp (~29% reduction in output current), Voperating becomes ~ 41 volts (NOCT) [using Suntech's graph]:
• (41v/35v) * 24 volts Vmp-noct-kyocera = ~ 28 volts (convert Suntech's Vop/Iop, because it has an I/V/P graph, to Kyocera Vop/Iop).

While that is not a high charging voltage--at (14.0 volt equivalent for a 12 volt bank)--it will charge OK. Obviously, as the battery bank becomes more charged, the final static voltage from the array (at Iop=0 amps) becomes:
• (~44v/35v) * 24 volts Vmp-noct-kyocera = ~30 volts (basically Kyocera's Voc-noct)

If they were put the array in series and install an MPPT charge controller, they could get probably upwards of 40% more power from their system (as a rough guess)...

-Bill
Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
• Solar Expert Posts: 336 ✭✭✭
Options
Re: Another weird setup

Thanks Bill, that makes sense.

Does the charge controller, therefore, in practice actually dictate the voltage to the panels?
Options
Re: Another weird setup

More or less--By the time you take heat depressing the Vmp of the solar array, voltage drop through the wiring and controller, and final charging voltage of the battery bank--Roughly Vmp~17.5 volts is the "optimum" for a PWM controller charging a 12 volt batter bank (assuming ~Vbatt-charge~14.5 volts or so). If you do all of the calculations, a hot array and cold battery bank will probably have difficulty reaching full equalization voltage--But that is probably only a limitation for the hot summer days in warm climates.

Remember, some PWM controllers can operate at higher voltages (Xantrex C40 is one)--However, your overall efficiency becomes (very roughly) 17.5 volts/Vmp-array (for a 12 volt array).

So a solar array with 40 volt Vmp charging a 12 volt battery bank on a PWM controller will have an efficiency of:

• 17.5 volts / 40 volts (array) (optimum Vmp-stc for 12 volt battery charging) = 0.4375 = ~ 44% charging efficiency with "high Vmp voltage" Array and 12 volt battery on PWM controller

For MPPT, assume the same basic starting/minimum array voltage (Vmp~17.5 volts @ STC), and then the controller design allows ~40-150 VDC maximum input voltage (Voc--open circuit voltage) for most controllers... A few vendors may offer 200+ Voc operation options/configurations (Midnite Classic, Xantrex XW MPPT 600 VDC, Apollo, etc.)...

In some cases, there are also MPPT controller limitations... For example, I believe, The Outback series has an "optimum" efficiency of Vmp/Vbatt of approximately 2:1 .... If you have a Vmp-array~100 VDC and a 12 volt battery bank, the controller will tend to run warm (is that conversion efficiency issues--or just normal I2R drop and diode/switching losses--Not really sure, I don't know enough about the controller).

Check the efficiency curves for the controllers you are thinking about using... Some have better efficiency than others. And some vendors do not publish efficiency graphs, so you may need to ask.

-Bill
Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
• Solar Expert Posts: 1,571 ✭✭
Options
Re: Another weird setup

It could be that the panels are connected in series with 52Vmp ? That'll definitely charge the battery, just very wasteful of panel power.