diy solar

diy solar

Shade question

M

MarkSolar

Solar Enthusiast
Joined
Mar 3, 2021
Messages
384
Location
Chicago far metro
I have 3 strings, each string 8 panels, arranged horizontally and combined in parallel to go into one tracker on a Fronius. Total for all panels is about 10kW. In the morning the first hour or so is shaded and I don't understand why the bypass diodes don't allow for the completely shaded panels to be bypassed. Here's a picture with about 1/3 of the system shaded and output 450W:
1003230838.jpg

Here it is with less shading and output still about 450W:

1003230844_HDR.jpg

And with only the last panel in each string shaded, the output is only up to 550W:
1003230856_HDR.jpg

But almost the instant the full sun hits those last panels, output goes up to 3.1kW, which is full output for 9am in the morning and then it rises quickly as the sun gets higher.
So my question is why don't the bypass diodes allow the entire shaded panels to be bypassed, which should allow the non shaded panels to send full output to the inverter? If I added an additional bypass diode to each of those shaded panels, would that allow the output to be higher in the morning?
 
First: Shade is devastating. You're lucky you're getting what you're getting.

Worth noting that while the shaded panels will not output much current, they will output near rated voltage.

Parallel into one tracker is likely the issue. ALL strings MUST operate at the same voltage. Bypass diodes reduce voltage by eliminating non-productive portions of the panels. Dropping the voltage of the non productive strings likely also reduces the output of the productive strings.

Have you considered re-arranging your strings? Since the shade moves laterally, if you ran your strings as follows:

1696348427626.png


The red and green strings would become fully un-shaded sooner thus increasing your production for those segments.

Even better option... cut down those effing trees.
 
sunshine_eggo said:
First: Shade is devastating. You're lucky you're getting what you're getting.

Worth noting that while the shaded panels will not output much current, they will output near rated voltage.

Parallel into one tracker is likely the issue. ALL strings MUST operate at the same voltage. Bypass diodes reduce voltage by eliminating non-productive portions of the panels. Dropping the voltage of the non productive strings likely also reduces the output of the productive strings.

Have you considered re-arranging your strings? Since the shade moves laterally, if you ran your strings as follows:
Click to expand...
Yes, if I'd known then what I know now I would have run the west half into one tracker and the east half into the other tracker because I have the same problem in the afternoon with tree shading. To change things now I'd have to run another set of wires to the inverter and the conduit isn't quite big enough. My need for electricity would be over if I cut down the trees because my wife would kill me.

I'm more curious at this point about how things work. This is my generic mental picture of a solar panel and the bypass diodes:

1696354637678.png

So let's just say Cell 3 in the diagram is one of my completely shaded panels at the end of the string, why aren't Cells 1 and 2 able to send their full output to the inverter?
 
MarkSolar said:
Yes, if I'd known then what I know now I would have run the west half into one tracker and the east half into the other tracker because I have the same problem in the afternoon with tree shading. To change things now I'd have to run another set of wires to the inverter and the conduit isn't quite big enough. My need for electricity would be over if I cut down the trees because my wife would kill me.
Click to expand...

LOL. Fair.

MarkSolar said:
I'm more curious at this point about how things work. This is my generic mental picture of a solar panel and the bypass diodes:

View attachment 170591

So let's just say Cell 3 in the diagram is one of my completely shaded panels at the end of the string, why aren't Cells 1 and 2 able to send their full output to the inverter?
Click to expand...

The MPPT has to be smart enough to identify the maximum power point, and the maximum power point may not be readily obvious. There can also be multiple local maximums, AND the MPPT has to conduct a sweep frequently enough to identify them.

cell 3 is cut out of the circuit, it's voltage is cut out, so the series string only has 2/3 the voltage of the series string.

Here's what my Victron does:

1696379908656.png

Those little dips are MPPT sweeps every 10 minutes. Note how they only dip down so far. Here's what voltage and current are doing:

1696380127126.png

Even with frequent sweeps, the MPPT isn't ever dropping below about 50% the Vmp. For you to see benefit, your MPPT would need to go to 1/12 the Vmp in 1/12th increments - likely below the functional voltage of the inverter.

Lastly, also worth noting that MPPTs are notoriously ineffective at identifying the maximum power point when the total insolation is lower.. like in the mornings and evenings.
 
MarkSolar said:
Yes, if I'd known then what I know now I would have run the west half into one tracker and the east half into the other tracker because I have the same problem in the afternoon with tree shading. To change things now I'd have to run another set of wires to the inverter and the conduit isn't quite big enough. My need for electricity would be over if I cut down the trees because my wife would kill me.

I'm more curious at this point about how things work. This is my generic mental picture of a solar panel and the bypass diodes:

View attachment 170591

So let's just say Cell 3 in the diagram is one of my completely shaded panels at the end of the string, why aren't Cells 1 and 2 able to send their full output to the inverter?
Click to expand...
They can send it, but your panels won't go into bypass mode if another string is putting out a higher voltage. And a lot of converters don't track beyond a local power point. Your controller doesn't even know those panels exist. Bypass diodes don't make up for bad design.
 
sunshine_eggo said:
LOL. Fair.



The MPPT has to be smart enough to identify the maximum power point, and the maximum power point may not be readily obvious. There can also be multiple local maximums, AND the MPPT has to conduct a sweep frequently enough to identify them.

cell 3 is cut out of the circuit, it's voltage is cut out, so the series string only has 2/3 the voltage of the series string.

Here's what my Victron does:

View attachment 170635

Those little dips are MPPT sweeps every 10 minutes. Note how they only dip down so far. Here's what voltage and current are doing:

View attachment 170636

Even with frequent sweeps, the MPPT isn't ever dropping below about 50% the Vmp. For you to see benefit, your MPPT would need to go to 1/12 the Vmp in 1/12th increments - likely below the functional voltage of the inverter.

Lastly, also worth noting that MPPTs are notoriously ineffective at identifying the maximum power point when the total insolation is lower.. like in the mornings and evenings.
Click to expand...
Thanks for the info and explanation. I'm actually very happy with how things turned out, it's paying for itself faster than I had calculated. The install I have is 1/2 our power, if I ever add the second half I'll configure things different, and might go for optimizers. I went for simple because I knew nothing, it's been a learning experience.
 
Bypass diodes or not, when two panels had a little bit of shade and one full string of 8 panels had full sun, no shade, should have output 1/3 of 3000W, for 1000W. But only if in full sun.

Shade appears because sun is at an angle, casting shadow from trees.
Sun at an angle will create less power than sun squarely on the panels.

By disconnecting 2 strings (shut off system for zero current before unplugging or pulling a fuse, OK to open a DC rated breaker with power on), you can test the output of each individually.

Can you log power all day long? You seem to indicate power jumps up/down the moment shade touches one panel. If that is true, something isn't working right. But I'm guessing it makes a nice smooth curve over the day.
 
Hedges said:
Can you log power all day long? You seem to indicate power jumps up/down the moment shade touches one panel. If that is true, something isn't working right. But I'm guessing it makes a nice smooth curve over the day.
Click to expand...
This is the plot from yesterday, when I took those photos. The first photo was at 8:38, second at 8:44 and third at 8:56. The sharp increase in slope on this plot begins at 8:45.
1696437866882.png\

I think if the trees weren't there those low shoulders at the beginning and end of day would be much steeper lines. I have done other experiments with shade. If I randomly hold an umbrella over some panel in various locations, it has little or no measurable effect. I've never tried throwing a tarp over an entire panel in one of the strings.
 
MarkSolar said:
To change things now I'd have to run another set of wires to the inverter and the conduit isn't quite big enough.
Click to expand...
This doesn't make sense. You currently have 3 strings that (i think) you combine into parallel and then enter the conduit. If you simply group the panels differently as @sunshine_eggo suggests into 3 strings, then combine into parallel, the same wires go thru the conduit.

Granted you may need to add some extensions (or possibly cut some shorter). But wither way you should have 8S3P and from where you connect in parallel, everything will be the same (except output likely higher).
 
Maybe as you say, without shade it would ramp up sooner, ramp down later. I think inverter will completely shrug off two parallel strings with some shade when one has no shade. And I think if diodes work it will bypass the shaded panel. Bad inverters might not, but Fronius is good.

An experiment with cardboard over a panel or two of a string could let you maintain the shaded condition for another hour, then suddenly switch to full sun. Then shade again. For an A-B test that doesn't also have changes in light intensity.

"But almost the instant the full sun hits those last panels, output goes up to 3.1kW"

Out of 10kW.

What is MPPT voltage range of inverter?
What is Voc & Vmp of your 8s strings?
I know some large Fronius are 1000V rather than just 600V inverters.
Some of my SMA inverters require significant voltage to deliver, while others can work over a wide range.
 
We're about at the equinox now where the sun is in the midpoint of the journey. If this is a new build, I would expect the shading to get worst in the winter, and better in the summer.

My particular shaded area I set my 410 watts of production, I will see 3 kWh in summer in the longest days where shade is no factor to as low as 1.5 kWh per day when the sun lower in the sky and shade from the trees is a factor.
 
chrisski said:
We're about at the equinox now where the sun is in the midpoint of the journey. If this is a new build, I would expect the shading to get worst in the winter, and better in the summer.

My particular shaded area I set my 410 watts of production, I will see 3 kWh in summer in the longest days where shade is no factor to as low as 1.5 kWh per day when the sun lower in the sky and shade from the trees is a factor.
Click to expand...
These are willow trees so they lose their leaves in winter. Also they're so tall that as the sun gets lower in the sky, it actually misses the canopy of the tree and there ends up being quite a bit less shade in the winter. I've had this 3 years, so I've got some historical data.
 
MisterSandals said:
This doesn't make sense. You currently have 3 strings that (i think) you combine into parallel and then enter the conduit. If you simply group the panels differently as @sunshine_eggo suggests into 3 strings, then combine into parallel, the same wires go thru the conduit.

Granted you may need to add some extensions (or possibly cut some shorter). But wither way you should have 8S3P and from where you connect in parallel, everything will be the same (except output likely higher).
Click to expand...
Yes that's correct if I used the approach @sunshine_eggo mentioned. And that's much less work than what I was referring to in regards to the conduit, which is to put the west half of the panels on one MPPT, the east half on the other. That would require running 2 sets of wires through the existing conduit that currently has one set of wires carrying the parallel combined strings.
 
You must log in or register to reply here.

Similar threads

L
Half-cut solar panel and shading
Replies
0
Views
140
LionKing
L
S
Adding to a string with shading
Replies
1
Views
111
Hedges
Hedges
P
Progressive Shading and Single MPPT
Replies
3
Views
113
pfowen
P
A
Single or Dual MPPT for Sliding Stacked Solar Panel Setup
Replies
1
Views
157
timselectric
T
A
Sunpower Performance 6 Panel Bypass diodes internal connections
Replies
1
Views
251
aprotopa
A
Back
Top