Shading effect question

[GaryGary]

Hi,
I recently finished a grid-tied system that uses ten 215 watt panels with the Enphase micro inverters.

This page from my website shows a situation in which 2 of the 10 panels get shaded early on some mornings by a fence post. The Enphase inverters appear to handle this well -- basically the two shaded panels produce part power, and the remaining 8 panels produce full power -- just like the Enphase blurbs say they should:
http://www.builditsolar.com/Projects...PV/Shading.htm

My question is what would the actual effect be if the panels were hooked up in a series string and connected to a single larger inverter. I've heard answers that range from little effect all the way to all of the panels would be knocked down to what the shaded panels are producing. Would really like to know how much the micro inverters are helping things (or not) in this kind of case.
I've Googled around for a good quantitative answer, and just not found anything.


Pages showing how the whole system is installed:
http://www.builditsolar.com/Projects...asePV/Main.htm

Thanks -- Gary

[RCinFLA]

The answer depends on the GT inverter minimum input voltage, how many bypass diodes are in a given panel, and how much voltage the series panel stack is normally operating at.

Can be anything from total shut down to losing just a portion of a shaded panel.

[BB.]

The "ideal" central (one) inverter would be to have all 10 panels in series connected to a 2+kW GT inverter.

Assuming that the minimum operational voltage of the inverter would support 8 panels in series ( Vmp array > Inverter low voltage)--then it should "gracefully" support one to two panel shading...

You could get into issues where you may "lose" 1/2 a panel's worth of energy from shading on the Enphase vs a "whole" panel with the inverter string (probably similar with the Enphase too--Like RCinFLA says, depends on how the shade falls and how the diodes are arranged).

If, the GT single inverter's low string voltage requires 10 panels minimum--then any shading of one to two panels will potentially "kill" the GT inverter's output until the shading is resolved. So--It depends on a lot of different issues (temperature of array, ambient temperatures too).

-Bill

[Dave Sparks]

I am playing around with high voltage strings right now and I think if you have this shading problem, you have to just test and try one or two of the "little boxes" and hope they help. Each make/model of panel has the potential to be fine without the "little box", need the "little box", or do nothing either way! Avoid shading ! Pretty soon this will not just be a grid-inverter problem but charge controller also. boB's little box should be out soon and something else is cooking also.

[GaryGary]

Thanks.

Sounds like there is not really a lot of difference unless the shading is enough to to drop the PV array voltage below the inverter minimum voltage.

Not to drag this out, but of the series connected array is partially shaded, but within the inverters allowed voltage range, will the inverters MPPT tracking be thrown off at all by having to look at a combination of shaded and unshaded panels?
Even if this were to happen, the MPPT being off a bit would not result in a big drop in power output from the unshaded panels?

Gary

[BB.]

It is difficult to explain--Solar panels are sort of the exact opposite of lead acid storage batteries (in an electronics sense).

A storage battery, basically tries to hold a constant voltage across its terminals until the current is too great (if discharging, the voltage will collapse; if charging, the voltage will rise dramatically). This is known as a "voltage source" in electronics.

Solar panels are "current sources"--basically, as long as the panel is between 0 volts and Vmp--The output current is pretty much fixed. (of course, there has to be a minimum amount of direct sunlight for any useful current flow--in shade, under heavy clouds, there is very little current flow).

Now, look at the power equation:

• Power = Voltage * Current (at least for DC)

Notice, for a storage battery, the voltage is (more or less fixed) from zero amps until the voltage collapse (too much current).

For a solar panel, as long as the current is less than Imp (current maximum power) for a particular amount of sun light (~1,000 watts per sq. meter is defined as "full sun) the voltage is kept relatively high--and you generate useful power (P=V*I)

If you are drawing too much current, voltage goes to near zero (voltage collapse)--and P=Imp*Zero.volts=zero.power (even though you have lots of current flowing).

And if you draw too little current, the voltage goes over Vmp towards Voc (voltage open circuit) you will hit Voc (maximum voltage) but the solar panel output approaches zero current--again zero power.

Here is a nice set of graphs for a generic solar panel:

• data sheet for a crystalline silicon panel

MPPT controllers attempt to find the sweet spot where Pmp=Vmp*Imp

When you have a shaded panel--the output current will fall with the degree of shading (sort-of--depends on the shape of the shadow, location of cells, bypass diodes, etc.)... But, once the current begins falling--that panel can no longer sustain 100% of Imp (remember, a solar panel's current is pretty much proportional to the amount of sunlight hitting the panel). So, that panel is no longer able sustain the current of the rest of the string--therefore it will "reverse bias" and the bypass diodes will pall the full current around the "dark cells/string of cells").

There are just so many variables--I don't think it is possible to give a generic answer... It is just better to assume that a partially shaded panel will dramatically drop that panel's output energy. And if that panel is in a series high voltage string (more than one panel in series)--if Vmp of the Array drops below Vin-solar of the charge controller/GT inverter--then that whole series string will not deliver any useful power into the solar device.

Even "soft shadows" (like a high voltage power line shadow across a solar array) can easily cause double digit (percentage) power losses.

My panels are on a second story roof--and I have just not bothered to try shading experiments on them--it was better to cut down/trim back the trees that where causing the shadows (trees cut down where too close to the foundation, bad roots, etc.).

Now, Vmp/Imp are not "sharp corners" in the power curve--and for a little bit of shading, it is possible that the results are not as "black and white" as I describe them--but avoiding shadows on solar PV arrays is critical for a good installation (move that vent pipe, position array away from chimney, etc.).

At least that is how I understand the problem...

-Bill

[Dave Sparks]

Bill,
Very well said! To the OP if the MPPT voltage is thrown off by some shade the diodes internal to the panels may still give you good output power. But they might not also. Each make and model of panel and the specific shading situation can be different. The "little black box" may or may not help. Too many variables. At some point you have to do the test or avoid shading at all cost!

[GaryGary]

Bill,
Very well said! To the OP if the MPPT voltage is thrown off by some shade the diodes internal to the panels may still give you good output power. But they might not also. Each make and model of panel and the specific shading situation can be different. The "little black box" may or may not help. Too many variables. At some point you have to do the test or avoid shading at all cost!

Hi,

Thanks.

I actually have the Enphase micro inverters (one per PV panel) on my system. I get just a little shading in the very early morning. The Enphase inverters appear to handle this well:
http://www.builditsolar.com/Projects...PV/Shading.htm

I was just wondering whether I should be congratulating myself on spending a few extra bucks for the Enphase inverters over a single string inverter, or be thinking I paid extra for something that ends up having little benefit :)
Sounds like "it depends".

The Enphase inverters did make the system very easy to install, so that's worth something in itself.

Gary

[BB.]

In the end, be happy with the choices you have made... There is nothing inherently wrong with either choice--just lots of opinions about what things may look like 15+ years ago--From Now .

My crystal ball is cracked.

-Bill

[niel]

15+ yrs ago? don't you mean 15+ yrs from now?