Why not mix panels?

andyrud
andyrud Solar Expert Posts: 70 ✭✭
Don't bite my head off!

In a recent thread someone had various sized panels and was told if they were within 5 or 10% of each other that would be OK. I read this all the time. BUT, in order to understand why, could you please explain what would happen. Would the strongest one just lower to the weakest one? Or would there be a fire? Or cause a panel to short out? Would DOGS AND CATS start living together!! What?

Thanks in advance,

Andy

Comments

  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: Why not mix panels?

    Andy;

    It's a matter of power loss. Let's look at two dissimilar panels: Kyocera 135 with a Vmp of 17.7 and Imp of 7.63 and a Kyocera 185 with a Vmp of 23.6 and Imp of 7.84.

    Panels are current sources, not Voltage sources. Normally you'd expect the panels to output a maximum of Vmp * Imp (17.7 * 7.63 = 135.051, 23.6 * 7.84 = 185.024). The trouble is with the two panels in parallel the Vmp will be limited by the lower panel. This means the larger panel essentially put out: 17.7 * 7.84 = 138.768. That's 46.256 Watts gone. Connect them in series and the Imp becomes the controlling factor and you only lose about 5 Watts.

    This is not a precise explanation, as there are many factors that affect a panel's actual output and which would come in to play in the power loss.
  • jaggedben
    jaggedben Solar Expert Posts: 230 ✭✭
    Re: Why not mix panels?
    The trouble is with the two panels in parallel the Vmp will be limited by the lower panel.

    Actually, the voltages will roughly average. But this does still mean that you will lose power from both panels, since neither will be operating at the optimal power point on the voltage*current curve.
    Connect them in series and the Imp becomes the controlling factor and you only lose about 5 Watts.

    I find this an odd calculation, because assuming the cells are actually identical in the panels you mentioned (and I believe they would be), there should be no loss of power in such a setup. You are just connecting 84 Kyocera polysilicon cells in series. I'm not sure why the ratio of Imp to number of cells should be any different for the two panels. (The ratio of Vmp to number of cells is identical.)

    Now, if you connect cells in series that definitely have different PV characteristics, then the characteristics of one panel may limit the output of the series string. That would be the case if you mixed, say, these Kyoceras with Suntech monosilicon panels. Most manufacturers don't provide enough detail about their cells such that ordinary folks could know if they can be mixed in series without power losses. And most panel combinations would result in power losses of some amount.
    This is not a precise explanation, as there are many factors that affect a panel's actual output and which would come in to play in the power loss.

    Hopefully I'm adding some precision, not confusing things. :blush:
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Why not mix panels?

    The physics of solar cells are actually pretty similar among brands and manfacturers (obviously there are differences in price, reliability, ultimate efficiency, etc.--but lets ignore those for now).

    Take a look at a generic solar panel specification (PDF download for a 245 watt panel).

    This vendor also supplied the Power Curve as well as the IV curve. If you look at the power curve, it is not a very sharp peak, but more of a rounded curve where being a few volts +/- of Vmp/Imp/Power peak does not really affect the over all panel output.

    That +/- 10% (or 5%) is based on that "soft" curve where you will not lose much energy from being less than perfect.

    However, you also have to look at the type of Charge Controller you are using.

    PWM controllers are (roughly) best matched with Vmp in the 17.5-18.X volt range to a 12 volt battery bank. Both in terms of efficiency and in terms of operational temperatures (hot panels, cold batteries--Vmp falls with increasing temperatures; Vbatt-charging increases with lower battery temperatures).

    So Vmp-hot just needs to be high enough to give Imp into a Vbatt-charging (with some losses for wiring and controller voltage drop).

    Note, since a PWM controller is just an "on / off" switch--Current in equals current out... so connecting a 35 volt Vmp panel to a 12 volt battery bank--Imp will not be much different...

    However, because Power = Voltage * Current... The excessively high Vmp of a miss-matched panel just "wastes" power... For example a Vmp=17.5 volt 5 amp panel vs a Vmp=35 volt at 5 amp:
    • Vmp*Imp = 17.5 volts * 5 amp = 87.5 Watt panel
    • Vbatt*Imp = 14.5 volts * 5 amps = 72.5 Watts into battery
    • Vmp*Imp = 25 watts * 5 amp = 175 Watt panel
    • Vbatt*Imp = 14.5 volts * 5 amps = 72.5 Watts into battery
    So, with a PWM type controller, the second panel will work, it just will not supply anywhere near its rated power.

    You do also have to watch for a Vmp=35 volt panel on some 12 volt controllers... The Voc-cold may be too high and damage the PWM controller's electronics.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset