Series vs Paralell Strings In Partial Shading (Trees, etc)

[Ron Swanson]

Question for you all.

We all know customers who don't want to cut down trees.  So we have patches of shade moving across various panels in an array.

We have 3 situations.  Let's say we have 6 panels.

1. All panels in series.  1000W/m2.  Shade moves across one panel bringing it down to 500W/m2.  8 amps generating in the rest.  4 amps generating in the partially shaded one.  The bypass diode is carrying the other 4.  Is the voltage of the partially shaded one now 0?  36V*8A*5 panels = 1440 watts?

2. All panels in parallel.  One half shaded.  [ 8A*5 panels + 4A*1 panel ] * 36V = 1584 watts?

3. Series/parallel system into combiner box?  One string at 108V nominal, the other at 72V nominal, MPPT controller drops voltage to 72 to get power from both strings, 1152 watts?  (String 1 108V*8A, string 2 72V*8A), 72V*16A = 1152W

Anybody ever done this testing in real life?  Am I understanding correctly?  And what is the configuration of the bypass diode in the panel?  Does it not take out the whole panel when some of the cells are shaded out?

[boB]

Hi Ron...  Tough subject to be totally accurate on all counts but will give it a try...

Question for you all.

We all know customers who don't want to cut down trees.  So we have patches of shade moving across various panels in an array.

We have 3 situations.  Let's say we have 6 panels.

1. All panels in series.  1000W/m2.  Shade moves across one panel bringing it down to 500W/m2.  8 amps generating in the rest.  4 amps generating in the partially shaded one.  The bypass diode is carrying the other 4.  Is the voltage of the partially shaded one now 0?  36V*8A*5 panels = 1440 watts?

The answer depends on how many bypass diodes are inside the panel and what sections of the module are shaded.  Theoretically if
you shade just ONE of those sections that are in series with the other sections, you will only lose that one section.  Each section
would act like its own solar panel and you would lose that amount of voltage and power.  Series  current would be the same.
That should be true until the voltage drops down to the battery voltage or lower, then the power output will drop drastically.
Some PV modules have 3 sections.

2. All panels in parallel.  One half shaded.  [ 8A*5 panels + 4A*1 panel ] * 36V = 1584 watts?

Again, depends on the structure of that one half shaded module and how it is shaded. 

BUT, in this case, you can pretty much kiss that modules contribution goodbye because its
MPP voltage will be below the others. While the MPPT controller or inverter should find
the best voltage, that voltage may just drop down very slightly below the normal MPPV
but won't pick up much of the capable output of the partially shaded module.

An optimizer ~might~ be able to boost the remaining output up to that of the other modules and recover
the rest of that modules' unshaded power.

3. Series/parallel system into combiner box?  One string at 108V nominal, the other at 72V nominal, MPPT controller drops voltage to 72 to get power from both strings, 1152 watts?  (String 1 108V*8A, string 2 72V*8A), 72V*16A = 1152W

The controller will run slightly above 72V MPPV to pick up some of the 108V string's power which is much higher than at 72 volts.
It probably won't go too much above the 72V though because the high side of the  MPP peak falls off  faster than below the peak.

This type of connection would lose a lot of power over being connected in series or all in parallel.

Anybody ever done this testing in real life?  Am I understanding correctly?  And what is the configuration of the bypass diode in the panel?  Does it not take out the whole panel when some of the cells are shaded out?

Again, it depends on the module's construction and how it is being shaded.   Some older modules (12V modules ?) had single
bypass diodes.  My old Kyocera 12V nominal PV modules had two 6V sections and each one had its own diode.  I don't know
how 12V modules are built these days.

Yes, these types of connections are fairly simple to predict if the shading area and module construction is known.

Best advice is to try and avoid shading best as you can or place the modules further away if possible and run wires
to the equipment.


boB

[Ron Swanson]

Of course the ideal situation would be for each module to have its own charge controller, if that were possible.

Are you aware of anyone who opted to break their array down into something like 2 kids instead of a single Classic?

[Sunshine]

The problem with breaking it down to two KIDs instead of one Classic is the two KIDs are still not as powerful, although it might still work out to more output in the long run because of the shading. That would be tricky to calculate. Alt E Store has a great video that explains shading in a simple way that most users can understand. boB has given you a more in-depth explanation but if you just want a customer to get the basics this covers it really well with good visuals. https://www.youtube.com/watch?v=ofo1HQyGG8s

[Ron Swanson]

Bob's explanation and that video were both great.  Seems like it pretty much lines up exactly with what I thought should happen.  It is always nice when mathematical predictions come true.

Saw a customer who had about 3000 watts of panels installed.  Due to sharp shade from close-in trees, I have never witnessed it over a couple hundred watts in winter.  Even though the majority of the array is in full sunshine.

Uniformity of illumination seems to be the #1 priority, and #2 would be keeping the number of panels per string as low as practical.

[Westbranch]

My own experience was with an infinitesimal shadow from an Aspen tree branch in the fall.
It was almost invisible.  But the Classic was telling me that somewhere across 12 PVs, 4s x  3p there was a shadow line and had just been halved. It was soooo Faint I did  not believe it!!!   

Chain saw out and power back up, 50% loss is too much. And it would have been an daily seasonal  event

This leads into boBs comments about panel construction and Cell orientation on the PV...

I suspected there might be some shadows that might impact but not that much, so I placed them in landscape orientation (Long side horizontal). It's just that Tree Branches don't grow STRAIGHT...

How about a higher Array height?  It is an OPTION...