How much "Power" do you give up from shadows?

jasonhc73

Cat herder, and dog toy tosser.
Joined
Oct 1, 2019
Messages
1,640
Location
Wichita, Kansas
One-third of one panel is shaded by the roofline. 2s2p - 270 W panels.

One-third of one panel reduced output nearly 40%

2021-10-04 (3).png2021-10-04 (2).png



361 watts vs 595 watts

2021-10-04 (1).png

2021-10-04.png

2021-10-04 (4).png
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,446
Location
Los Gatos CA
Isn't it amazing how just a little shade can cause so much loss!!

Are there blocking diodes on each serial string? A shaded string in parallel with a 'good' string can pull down both. A blocking diode can help prevent the shaded string from impacting the good string.

The panels probably already have bypass diodes but if they don't you should consider adding them. (With the parallel string, bypass diodes might not help much but it is cheap to try. )



I just looked closer at the pics. One of them shows an MPPT1 and an MPPT2 (unused). If there are two MPPT inputs, putting each string on a different MPPT is even better than blocking diodes.
 

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
Joined
Oct 29, 2019
Messages
5,408
Location
Rural NE Ontario Canada
Shading is handled by Solar Panels differently depending on their design & materials used, they are quite different actually and some are Far Better in Shaded Environments than others. Anyone considering buying Panels should consider the Type of Panel which is most suitable to their specific location. For example Split-Cell panels will perform much better in shade than others.

I Very Strongly suggest members read the following Two Articles:
 

jasonhc73

Cat herder, and dog toy tosser.
Joined
Oct 1, 2019
Messages
1,640
Location
Wichita, Kansas
Isn't it amazing how just a little shade can cause so much loss!!

Are there blocking diodes on each serial string? A shaded string in parallel with a 'good' string can pull down both. A blocking diode can help prevent the shaded string from impacting the good string.

The panels probably already have bypass diodes but if they don't you should consider adding them. (With the parallel string, bypass diodes might not help much but it is cheap to try. )



I just looked closer at the pics. One of them shows an MPPT1 and an MPPT2 (unused). If there are two MPPT inputs, putting each string on a different MPPT is even better than blocking diodes.
Just one MPPT, lv2424. Solar-assistant.io shows two MPPTs by default.
 

wattmatters

Solar Addict
Joined
Apr 16, 2021
Messages
852
Do you know if the panels have bypass diodes? (I am guessing they don't.)
They will.

The panels shown are 72 cell panels and would be configured into three groups of 24 cells, each making up two columns of 12 cells high x 2 cells wide. Each group of 24 cells will have a bypass diode.

Like this, except this image is a 60 cell panel with 10 cells in each column - the same principle applies.
bypass-diodes.jpg


This is why when covering the whole bottom row of the panel severely current limits the panel as each group of 24 cells are blocked. This sort of shade will murder a panel's output and can severely restrict a string depending on how well the diodes can bypass the panel entirely, not all can (this is where optimisers help).

In the case of single cell coverage and single column coverage, only one group of 24 cells is blocked and it's bypass diode then passes current around those cells enabling the other two sets of 24 cells (and the other 2 panels) to continue production as normal.

The losses shown in the video in both the single cell and single column shade cases were ~75W or ~11% of the unshaded 650W output.

Across three panels there are 9 sets of 24 cells. 1 set out of 9 sets = 11%.
 

wattmatters

Solar Addict
Joined
Apr 16, 2021
Messages
852
Not all panels are arranged in this manner, some have a half cut design which handle these bottom of panel shading scenarios better. These half cut panels essentially split the panel into two small panels with the upper and lower sections able to operate independently (but the overall drop in current will still hurt production).

If you have panels which will cop some form of unavoidable partial block shading (rather than say variable/intermittent shading), then it can help to think about whether the shade comes from the side, or from the bottom. This can inform whether a panel is better placed in portrait or landscape orientation.

Example of panel array placed on a sloped rooftop above where a brick chimney emerges:

IMG_2724.jpeg
While in this image the shadow is not on the panels, as the seasons move towards Winter and the Sun gets lower in the sky then the chimney casts block shade over the lower row of cells on one or two panels. This small amount of block shading can significantly hurt production from that array.

Similar block shade entering the panel from the side would not be as limiting as the bypass diode can help, at least until it casts shadow over most of the panel. These scenarios are more likely to occur at the start and end of the day when production is low anyway, so the absolute losses are not so bad.

But block shading covering the lower rows tends to be a seasonal issue even during the middle of the day when production should be highest and it can really hurt potential Winter production.
 

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,446
Location
Los Gatos CA
They will.

The panels shown are 72 cell panels and would be configured into three groups of 24 cells, each making up two columns of 12 cells high x 2 cells wide. Each group of 24 cells will have a bypass diode.

Like this, except this image is a 60 cell panel with 10 cells in each column - the same principle applies.
bypass-diodes.jpg


This is why when covering the whole bottom row of the panel severely current limits the panel as each group of 24 cells are blocked. This sort of shade will murder a panel's output and can severely restrict a string depending on how well the diodes can bypass the panel entirely, not all can (this is where optimisers help).

In the case of single cell coverage and single column coverage, only one group of 24 cells is blocked and it's bypass diode then passes current around those cells enabling the other two sets of 24 cells (and the other 2 panels) to continue production as normal.

The losses shown in the video in both the single cell and single column shade cases were ~75W or ~11% of the unshaded 650W output.

Across three panels there are 9 sets of 24 cells. 1 set out of 9 sets = 11%.
I had to go watch the video again but I see it now. The first time I watched it I was reversing the rows and columns in my mind so so I misinterpreted what was going on. Too bad he did not talk about the bypass diodes in the video. If he did it would be a great video to show the benefits of bypass diodes.

Also, it was hard to see, but when there was no shade, it was 95V and when one cell was shaded, it was 84.5 volts. This fits perfectly with one of the zones on one of the 3 panels being bypassed. There are a total of 9 zones in series so 8/9 x 95 = 84.4

Just as an aside, those 3 diode panels are intended for mounting with the zone stripes being horizontal. That way, as the sun rises, it will start producing when only 1/3 of the cell gets full sun and then 2/3s and then the whole panel. The opposite happens in the eve. (This is probably why I confused the row and columns when I first watched the video)
 
Last edited:

FilterGuy

What, me worry?
Joined
Nov 26, 2019
Messages
4,446
Location
Los Gatos CA
Not all panels are arranged in this manner, some have a half cut design which handle these bottom of panel shading scenarios better. These half cut panels essentially split the panel into two small panels with the upper and lower sections able to operate independently (but the overall drop in current will still hurt production).

If you have panels which will cop some form of unavoidable partial block shading (rather than say variable/intermittent shading), then it can help to think about whether the shade comes from the side, or from the bottom. This can inform whether a panel is better placed in portrait or landscape orientation.

Example of panel array placed on a sloped rooftop above where a brick chimney emerges:

View attachment 67685
While in this image the shadow is not on the panels, as the seasons move towards Winter and the Sun gets lower in the sky then the chimney casts block shade over the lower row of cells on one or two panels. This small amount of block shading can significantly hurt production from that array.

Similar block shade entering the panel from the side would not be as limiting as the bypass diode can help, at least until it casts shadow over most of the panel. These scenarios are more likely to occur at the start and end of the day when production is low anyway, so the absolute losses are not so bad.

But block shading covering the lower rows tends to be a seasonal issue even during the middle of the day when production should be highest and it can really hurt potential Winter production.
Yes. The half-cut cell panels are something I have not dug into (I have not used them so I have not had a need).

As I understand it, the two zones of the half-cut cells are wired in parallel. Is that correct? Do you happen to know if there are blocking diodes between the two zones? This would help prevent the shaded half from pulling down the other half.

The other thing I wonder about the half-cut panels is what happens when 2 or more of these are in series and half of one is shaded. In this case, one panel will produce about 1/2 the current of the other..... I think that means it will cut the whole string current in half. (As I type this, I am wondering if some of the new panels with really high Voc (>60V) are half-cut with the 2 zones wired in series? )

All of this adds up to this: If you are going to have shading, it is important to know how your panels work and lay them out to minimize the impact of the shading.
 

fafrd

Photon Sorcerer
Joined
Aug 11, 2020
Messages
2,205
Yes. The half-cut cell panels are something I have not dug into (I have not used them so I have not had a need).

As I understand it, the two zones of the half-cut cells are wired in parallel. Is that correct? Do you happen to know if there are blocking diodes between the two zones? This would help prevent the shaded half from pulling down the other half.

The other thing I wonder about the half-cut panels is what happens when 2 or more of these are in series and half of one is shaded. In this case, one panel will produce about 1/2 the current of the other..... I think that means it will cut the whole string current in half. (As I type this, I am wondering if some of the new panels with really high Voc (>60V) are half-cut with the 2 zones wired in series? )

All of this adds up to this: If you are going to have shading, it is important to know how your panels work and lay them out to minimize the impact of the shading.
I’ve got half-cut panels.

They are two half-panels connected in parallel and sharing a common set of three bypass diodes.

So with just one panel, if one half panel has shade impacting 2/3 of the columns, that panel will operate at 50% max power (full Vmp with 1/2Imp).

If one half panel only has shade impacting 1/3rd of the columns, the bypass diode will activate for 67% max power (2/3Vnp with full Imp for both half-panels).

If there are 2 half-cut panels in parallel, shade completely blocking a single half-panel will result in 75% max power (3 out of 4 half-panels at full Vmp and 1/2Imp). No bypass diodes activated.

If 2/3 of the columns of a single half-panel are blocked by shade, the parallel string will continue to operate as above since activating 2 bypass diodes would reduce output power of the shaded panel as well as the shorted unshaded panel in parallel to 33% of max resulting in only 33% power for the full 2P string (less than that available by operating at full Vmp with only 3 out of 4 half panels contributing 1/2Imp).

And the same will be true if shade is blocking only 1/3 of one half-panel, a single bypass diodes activating would resulting in 67% full power from both the partially-shaded panel and the unshaded panel, less that the 75% available by operating at full Vmp with 1-1/2 Imp (so the MPPT controller will prefer to operate at Vmp for 75% max output).

Put those same 2 half-cut panels into a 2S series string and the impact of shade affecting a single half-panel will pretty be identical to what you’d get with standard (non-half-cut) panels.

If shade completely blocks one half-panel, total power is reduced to slightly more than 50%. Because only 1/2 Imp flows from the partially-shaded panel (with voltage of Vmp), the second unshaded panel is limited to that same 1/2 Imp of current. Because each half panel is only passing 1/2 Imp, voltage across the unshaded panel will increase above Vmp (hence the >1/2 Wmp power output from the second panel).

A 2S string of half-cut panels behaves like a standard panel if shading on one half-panel is only blocking 1/3 of the columns. In that case, 1 bypass diode will be activated delivering full Isc at 2/3Vmp from the partially-shaded panel meaning total 2S string voltage will be 5/6ths of what it would be without shade, or 83.3% (superior to the 75% of the 2P string).

And shade impacting 2/3 of one half panel will activate 2 bypass diodes, resulting n 67% of the maximum (unshaded) output for the 2S string (inferior to the 75% of the 2P string).

The main advantage of half cut panels in shade is that you can go massively parallel allowing all half-panels to operate at Vmp from a single MPPT getting full power from any unshaded half-panels in the full parallel string…
 
Top