Will shading keep me from getting ANY solar power in the winter?

[Sprucebeach]

I'm considering a grid-tied system that would also be off grid during our frequent power failures. I would anticipate putting a row of panels in portrait mode just below the peak of my roof (that faces 187°) and another row below them across the upper part of the front-porch roof - about 30 panels total. My house is in a clearing on a ridge heavily wooded with mostly deciduous trees and the roof is the only location where panels can be mounted. Tree cutting isn't an option for us, and trimming enough to get good sun in the winter would remove most of the blocking trees' vegetation.

For six months of the year both parts of the roof are in full sun for close to five hours a day. But as the year wears on the morning shadows are slower to sweep off the right edge of the roof and the afternoon shadows come in earlier from the left and inexorably advance higher and higher on the roof. Before the leaves fall the shade can be heavy. Below is the roof on Oct. 22 at 10:00, noon, and 1:00 p.m. (All times are corrected to Standard Time.)





After the leaves are mostly off the trees more light falls on the roof but it is mottled by the abundance of tree trunks and branches. Below is a series of shots from beginning of December.



Unfortunately my schedule didn't permit a noon shot but I think the picture is pretty clear without it: For the darkest three-or-so months of the year only small patches of uninterrupted sunlight would move across the panels. Most panels would see mottled sunlight most of the time.

I have two questions.

1. How much PV will I be able to get during these winter months? Can I get any at all?

2. Assuming there is PV to be gotten, what is the optimum way to get it? Microinverters? Optimizers? And how would I configure the string layout?

As you can probably tell from where this is posted, I'm a complete novice who is trying to educate himself so as to be able to make an informed decision about whether I should go solar. I'd welcome any comments.

 

[Dzl]

Damn, with the exception of 11am kinda, that looks really rough.

I don't know much about residential solar, but I think Microinverters or DC optimizers are the best bets for partial shade performance, along with maybe panels made from half cut cells with 2x the number of bypass diodes, but I think in all these cases they are meant to deal with bits and pieces of shade here and there (like a chimney or vent pipe or branch). I'm not sure how effective they could be with such extensive and spread out shading.

But like I said, I really don't know much when it comes to residential solar, and shading seems a complicated topic. Others will come along with better insight/more knowledge than me.

 

[fafrd]

I'm considering a grid-tied system that would also be off grid during our frequent power failures. I would anticipate putting a row of panels in portrait mode just below the peak of my roof (that faces 187°) and another row below them across the upper part of the front-porch roof - about 30 panels total. My house is in a clearing on a ridge heavily wooded with mostly deciduous trees and the roof is the only location where panels can be mounted. Tree cutting isn't an option for us, and trimming enough to get good sun in the winter would remove most of the blocking trees' vegetation.

For six months of the year both parts of the roof are in full sun for close to five hours a day. But as the year wears on the morning shadows are slower to sweep off the right edge of the roof and the afternoon shadows come in earlier from the left and inexorably advance higher and higher on the roof. Before the leaves fall the shade can be heavy. Below is the roof on Oct. 22 at 10:00, noon, and 1:00 p.m. (All times are corrected to Standard Time.)

View attachment 74400
View attachment 74401
View attachment 74403

After the leaves are mostly off the trees more light falls on the roof but it is mottled by the abundance of tree trunks and branches. Below is a series of shots from beginning of December.

View attachment 74389View attachment 74390View attachment 74395View attachment 74393View attachment 74394

Unfortunately my schedule didn't permit a noon shot but I think the picture is pretty clear without it: For the darkest three-or-so months of the year only small patches of uninterrupted sunlight would move across the panels. Most panels would see mottled sunlight most of the time.

I have two questions.

1. How much PV will I be able to get during these winter months? Can I get any at all?

2. Assuming there is PV to be gotten, what is the optimum way to get it? Microinverters? Optimizers? And how would I configure the string layout?

As you can probably tell from where this is posted, I'm a complete novice who is trying to educate himself so as to be able to make an informed decision about whether I should go solar. I'd welcome any comments.

I’ve got a very similar situation. I elected to go for a 1S parallel array of half-cut panels to maximize my output in the presence of partial shading.

Half-cut panels are essentially 2 half-panels in parallel, so you’ll get half-output from several panels at noon on your Oct 22 pic.

By connecting all panels in parallel, you’ll get the maximum current that each half-panel can contribute without any shaded panel blocking or starving any other panel (and without needing to purchase any optimizers).

A 1S array means higher currents, so you’ll need to invest in heavier-gauge home run wires if you are going for a DC-coupled array.

The other way to achieve the same output as a parallel 1S array is to go with a Microinverter-based array (still using half-cut panels) - this will allow each panel to contribute it’s maximum output current before converting it to AC power.

I was getting 8kWh from my 1.14kW 1S3P array this summer and am still getting ~1/3 of that amount today (December 1).

Interestingly, on overcast days I can actually get as much output as I get on clear days. The clouds reduce overall output levels but the diffused light eliminates the shading so I get about as much output as I’d get without any shading (in the presence of overcast / clouding).

But half cut panels are the only hope you have of salvaging much of any output with that extreme degree of shading.

What about the other roof orientations? Any better as far as shading?

 

[chrisski]

I’ve got a very similar situation. I elected to go for a 1S parallel array of half-cut panels to maximize my output in the presence of partial shading.

The panels I tested did not work good in shade. When I have 3 panels in series, if I shade two cells I lose between 1/4th and 1/3rd the output. When I have a single panel in series, if I shade two cells, I lose between 2/3 and 3/4 the output.

I tested two different 100 watt panels. I tested 100 watt Renogy and 100 watt Lion Energy with similar results. Shading two cells is less Than the branches on those leafless trees.

I would not want to put panels on that roof unless I don’t mind losing production several times a year. I do not have house panels, so I’m not the best judge.

I would think the OP may need to choose a panel based off being good in shade, but I have no idea where to find that list. Seems like this may be a candidate for CIGs, but those are really not available for house installations and are not that efficient.

 

[fafrd]

The panels I tested did not work good in shade. When I have 3 panels in series, if I shade two cells I lose between 1/4th and 1/3rd the output.

You’ve got to be more clear about exactly what you are shading.

3 panels in series with 1 cell fully shaded in two separate panels or 2 separate 1/3 segments of a single panel should result in over 1/3rd power loss. For example, if we assume Vmp of a single panel is 50 times the diode drop of a single bypass diode, string voltage will drop from Vmp to 62.8% Vmp (2/3 Vmp - 2 diode drops).

When I have a single panel in series, if I shade two cells, I lose between 2/3 and 3/4 the output.

If you’ve fully shaded a single cell on two separate conventional panels in series, output will drop by 2/3 if no bypass diodes get activated (poor MPPT controller).

A good MPPT controller will find that lowering voltage to 2/3 Vmp - 1 diode drop will decrease output loss to 35.3%.

I tested two different 100 watt panels. I tested 100 watt Renogy and 100 watt Lion Energy with similar results. Shading two cells is less Than the branches on those leafless trees.

I would not want to put panels on that roof unless I don’t mind losing production several times a year. I do not have house panels, so I’m not the best judge.

I would think the OP may need to choose a panel based off being good in shade, but I have no idea where to find that list. Seems like this may be a candidate for CIGs, but those are really not available for house installations and are not that efficient.

 

[chrisski]

3 panels in series with 1 cell fully shaded in two separate panels or 2 separate 1/3 segments of a single panel should result in over 1/3rd power loss.

This is a thread I did of the shading two cells on a single panel with three in series. I did this on two different strings. I did not do a thread on the single panel with two cells shaded.

Parallel Panels and Shading versus Series Parallel

I tested some panels for output today when shaded. 2 Panels in Parallel (36 Cells Per Panel) -One panels with only 2 cells shaded 26% Reduction in Wattage -One Panel 50% shaded 54% Reduction in Wattage 3 Panels in Series (33 Cells Per Panel) -One panel Two cells Shaded 33%-50% Reduction in...

diysolarforum.com

 

[fafrd]

This is a thread I did of the shading two cells on a single panel with three in series. I did this on two different strings. I did not do a thread on the single panel with two cells shaded.

Parallel Panels and Shading versus Series Parallel

I tested some panels for output today when shaded. 2 Panels in Parallel (36 Cells Per Panel) -One panels with only 2 cells shaded 26% Reduction in Wattage -One Panel 50% shaded 54% Reduction in Wattage 3 Panels in Series (33 Cells Per Panel) -One panel Two cells Shaded 33%-50% Reduction in...

diysolarforum.com

Your other post needs to provide a great deal more detail about exactly which cells were shaded for me to comment on whether your results make sense or not.

As far as shading only a single cell on two separate panels in a 3P array, any half-decent MPPT would activate a single bypass diode in each of those shaded panels by dropping string voltage to ~2/3Vmppt - Vdiode resulting in power loss of about ~1/3, so if you are losing 2/3 of your power, the only explanation is that you either have a PWM controller or a piss-poor MPPT controller that fails to find the increased output power available at ~2/3Vmppt and continues to drive the string at full Vmppt (meaning the 2 partially shaded panels are choked off to 0% output and only the single unshaded panel is producing power)…

 

[robby]

I wonder about the panels themselves. Several new tests have been done showing shading has a minimal effect on series panels and they have suggested that it effects are minimum on parallel as well.

 

[chrisski]

Your other post needs to provide a great deal more detail about exactly which cells were shaded for me to comment on whether your results make sense or not.

My point to this post is to not give the OP too much hope that shading like he pictures will produce a lot of power

Based off what I see on shading, if that entire south facing roof were covered in panels, I would expect them to produce less than 25% of rate output at any one given time. I’ve only done my shade tests with 100 watt panels. Perhaps roof panels are different.

Would you expect more output than that?

 

[TomC4306]

Several new tests have been done showing shading has a minimal effect on series panels

Please support this assertion with a link.

 

[Ceberha]

I have a similar situation with 310 watt panels each with individual enphase microinverters, ie 4 or 5 hours of sun in summer but dappled shade from surrounding deciduous trees in winter. In Nov to January the system puts out 10 to 20% of the June to August numbers. This is in the mid Atlantic region.

 

[robby]

Please support this assertion with a link.

Scientific controlled tests done by NRG

Another test showing the String vs Micro

This guy has been saying it for years.

I got interested in this subject when my own roof installer guy said this to me and I argued with him and he said he had tried it himself several times and found very little difference between Micro and String. He told me it's was a real issue a decade ago or if you buy older panels that do not have bypass diodes and also cheap or old MPPT trackers but he said today he see's almost no difference between the two when handling shading today. I kind of ignored his idea at first but then one day I looked it up and found several guys saying the same thing but it was only when NRG recently did the test scientifically that it left no doubt in my mind.

 

[fafrd]

My point to this post is to not give the OP too much hope that shading like he pictures will produce a lot of power

Based off what I see on shading, if that entire south facing roof were covered in panels, I would expect them to produce less than 25% of rate output at any one given time. I’ve only done my shade tests with 100 watt panels. Perhaps roof panels are different.

Would you expect more output than that?

Those thick tree branch shadows are bad, He might get close to 50% output from a few panels near the clearest part of the day but anything close to 25% full-day unshaded output for the full array would be an achievement.

But that’s only in winter (when the sun is low). During summer months when the sun is higher in the sky, he’d probably get at 3 to 4 times wintertime output…

50% unshaded output in summer (when ~2/3 out annual output is produced) plus 25% unshaded output in winter (when the remaining ~1/3 of unshaded output is produced) translates to over 40% of unshaded output annually.

So planning for an array 2.5x the size needed for target production without shade sounds realistic (though I still wonder about other roof faces).

 

[robby]

I agree that the OP is not going to get all that much power out of the system.

I think if he wants it purely for battery backup during a power outage then a relatively inexpensive inverter setup and a nice bank off AGM batteries would be the ideal solution.

If money is not an issue and he wants a whole house solution then an SMA Battery Backup with AGM batteries and some panels should do the job nicely.

 

[Sprucebeach]

Scientific controlled tests done by NRG

Another test showing the String vs Micro

This guy has been saying it for years.

I got interested in this subject when my own roof installer guy said this to me and I argued with him and he said he had tried it himself several times and found very little difference between Micro and String. He told me it's was a real issue a decade ago or if you buy older panels that do not have bypass diodes and also cheap or old MPPT trackers but he said today he see's almost no difference between the two when handling shading today. I kind of ignored his idea at first but then one day I looked it up and found several guys saying the same thing but it was only when NRG recently did the test scientifically that it left no doubt in my mind.

I sounds from the Solar Ray video that with string inverters there's not even a need for optimizers. My shading situation seems quite different from the shading in his testing though. And it's extremely different from what was tested in the NRG video, in which only a tiny fraction of the total panel surface area was being shaded. Does anyone have any feeling for whether optimizers would be useful in the case of my winter shading?

 

[Sprucebeach]

Those thick tree branch shadows are bad, He might get close to 50% output from a few panels near the clearest part of the day but anything close to 25% full-day unshaded output for the full array would be an achievement.

But that’s only in winter (when the sun is low). During summer months when the sun is higher in the sky, he’d probably get at 3 to 4 times wintertime output…

50% unshaded output in summer (when ~2/3 out annual output is produced) plus 25% unshaded output in winter (when the remaining ~1/3 of unshaded output is produced) translates to over 40% of unshaded output annually.

So planning for an array 2.5x the size needed for target production without shade sounds realistic (though I still wonder about other roof faces).

Thanks for these comments, which seem to support those of chrisski and Ceberha above, which I also very much appreciate. I'd like more output during the winter months but I can live with this kind of output reduction.

Regarding other roof faces -

The house is L-shaped and there is a shorter roof extending south at a right angle off the left side of the roof in the pictures. The pictures show the very beginning of the south-running roof. The west face of that roof gets almost zero sun because of close-by trees. The east face gets morning sun but only in the summer. Because the roof is extended to the south it's closer to more screening trees there, and during the winter it gets very little sun. So it doesn't help my winter problem, and in the summer I'd already be getting all the sun I needed from the main and porch roofs.

 

[fafrd]

I appreciate this helpful information. I was interested to see Ceberha's experience with a similar shading situation using Enphase microinverters. I could live with that reduction during the winter months.

I sounds from the Solar Ray video that with string inverters there's not even a need for optimizers. My shading situation seems quite different from the shading in his testing though. And it's extremely different from what was tested in the NRG video, in which only a tiny fraction of the total panel surface area was being shaded. Does anyone have any feeling for whether optimizers would be useful in the case of my winter shading?

When shading cuts across a full panel of a 1PxS string, optimizers won’t help you.

When shading blocks a single cell of a panel in a 1PxS string, optimizers won’t help you.

When a strange diffused shading situation partially shades one or more cells by 25-50% (or cuts down light intensity reaching multiple cells of one specific panel by 25-50%, optimizers will help you.

If you have 2 xS strings in parallel (2PxS array) feeding a single MPPT, optimizers can help you, but investing in a second MPPT would likely give you better bang for your buck).

Optimizers can’t do much to help with ‘hard shading’ fully blocking full cells in a 1PxS string.

Only in the case of diffused shading resulting in partially-degraded output current and/or 2 or more series strings feeding the same MPPT in parallel will they really make enough of a difference to justify their cost.

 

[robby]

I appreciate this helpful information. I was interested to see Ceberha's experience with a similar shading situation using Enphase microinverters. I could live with that reduction during the winter months.

I sounds from the Solar Ray video that with string inverters there's not even a need for optimizers. My shading situation seems quite different from the shading in his testing though. And it's extremely different from what was tested in the NRG video, in which only a tiny fraction of the total panel surface area was being shaded. Does anyone have any feeling for whether optimizers would be useful in the case of my winter shading?

Sorry for the confusion I was just answering @TomC4306 question about proof of shading having a minimal impact on string inverters. It's been a myth spread by the Micro Inverter companies that if even if a leaf gets on your panels your power will drop significantly. Your situation is not minor shading, it's major shading. You just cannot produce significant power with minimal sunlight.

My proposal of just having an AGM battery backup system should work. flooded Lead Acid and AGM batteries have no problem with remaining fully charged for long periods. So your minimal power will keep them charged for those days when you do have a power outage. The SMA Battery Backup unit will do an automatic transfer of your whole house to batteries if you have an outage. You could add in a SMA Sunny Boy Inverter if you want to grid tie and also have the solar power to your house, I just don't think your going to get a good return on the Investment based on the shading problem.

 

[Sprucebeach]

When shading cuts across a full panel of a 1PxS string, optimizers won’t help you.

When shading blocks a single cell of a panel in a 1PxS string, optimizers won’t help you.

When a strange diffused shading situation partially shades one or more cells by 25-50% (or cuts down light intensity reaching multiple cells of one specific panel by 25-50%, optimizers will help you.

If you have 2 xS strings in parallel (2PxS array) feeding a single MPPT, optimizers can help you, but investing in a second MPPT would likely give you better bang for your buck).

Optimizers can’t do much to help with ‘hard shading’ fully blocking full cells in a 1PxS string.

Only in the case of diffused shading resulting in partially-degraded output current and/or 2 or more series strings feeding the same MPPT in parallel will they really make enough of a difference to justify their cost.

Thanks for that very clear discussion.

 

[fafrd]

Sorry for the confusion I was just answering @TomC4306 question about proof of shading having a minimal impact on string inverters. It's been a myth spread by the Micro Inverter companies that if even if a leaf gets on your panels your power will drop significantly.

Uhhh, sorry, but that’s an overstatement (at least when half-cut panels are considered).

Let’s say a leaf blocks 2 adjacent cells under control of seperate bypass diodes by ~75%. With conventional panels, both shaded segments get bypassed reducing the shaded panels output by 66%.

With conventional panels in a very long series string, that’s a minor output drop and not much different than what you’d get with Microinverters.

But with a short string of 4S or 3S, it’s a drop of 17% to 22%. Microinverters are not going to do any better with conventional panels, but hard to call that level of production loss ‘minimal’.

Use half-cut panels in that same 4 or 3 panel array and a Microinverter will limit production loss to 12.5% (4S) to 17% (3S), so about 3/4 of the loss associated with either panel type in a series string.

Of course if you use half-cut panels in a long series string, they won’t make much difference in this shading scenario (since limiting the entire string to half-current is far worse than just dropping voltage of a single panel from full Vmp to 1/3Vmp.

But for modest-voltage strings composed of a relatively small number of half-cut panels in series, Microinverters absolutely provide superior output to string inverters.

Your situation is not minor shading, it's major shading. You just cannot produce significant power with minimal sunlight.

My proposal of just having an AGM battery backup system should work. flooded Lead Acid and AGM batteries have no problem with remaining fully charged for long periods. So your minimal power will keep them charged for those days when you do have a power outage. The SMA Battery Backup unit will do an automatic transfer of your whole house to batteries if you have an outage. You could add in a SMA Sunny Boy Inverter if you want to grid tie and also have the solar power to your house, I just don't think your going to get a good return on the Investment based on the shading problem.