diy solar

diy solar

Feedback on my DIY bifacial panel unistrut ground mount design?

AxelHeyst

New Member
Joined
Nov 24, 2022
Messages
13
Location
Southern Sierra Nevadas, CA
Hello, this is my first post, but have gained from this forum for some time now.

I'm helping a neighbor upgrade his system. He's getting (12) new bifacial panels. I'm designing a ground rack for him. If you'd care to, I'd love any feedback. See any problems with my design? Think it'll blow over? Gut check? Any feedback is much appreciated.
  • Snow loads are minimal. It'll snow a foot at most here, and then melt.
  • Wind can be stiff, but we have no hurricanes or tornadoes here.
  • All of the fittings are standard unistrut fittings.
  • Fitting attachment to the piers is deep jbolts embedded, not drilled expansion bolts.
  • The only non-unistrut stuff besides the concrete is the diagonal cross braces, which I'm thinking I'll use something like ~1-2"dia galvanized metal fence pipe.
Okay lets see if I can insert images correctly...
20230130_3x4_0010_converted.jpg20230130_3x4_0020_converted.jpg20230130_3x4_0030_converted.jpg20230130_3x4_0040_converted.jpg20230130_3x4_0050_converted.jpg20230130_3x4_0060_converted.jpg20230130_3x4_0070_converted.jpg20230130_3x4_0080_converted.jpg20230130_3x4_0090_converted.jpg20230130_3x4_0100_converted.jpg
 
What is the latitude of the installation? The angle seems steep.

How will the rack be grounded? Will there be WEEB washers to electrically bond everything together?
 
Latitude is 35* (southern Sierra Nevada mountains, California). We're overpaneled for summer months, and winter demand includes a heat pump and other equipment during the winter, so I'm biasing towards winter production. Cloudy days are infrequent here; multiple cloudy days in a row exceptionally rare. There is generator backup.

Additionally, I had a hard time finding fittings that would allow me to run a more traditional 35* tilt angle. They were either very expensive, or unavailable from the outfitters I could find.

I haven't designed the grounding method yet, so the short answer is "it will be grounded appropriately". Thank you for the link! I will review it carefully. I have another build that I need to ground the panels on so it's very timely.
 
These are the graphs for system SoC comparing effect of tilt angle. I was surprised the effect of tilt wasn't more. (I'm assuming backfacing production is zero here.)
The winter demand I'm showing here it worst realistic case (I believe), running the heat pump 24hrs for heating. I wouldn't want to see 50% DoD every day.
summerschedule.pngwinterschedule.png
 
Hey @AxelHeyst, that looks good. The level ground looks especially nice ;-)

Depending on wind load, you can probably just use a couple of the 90 degree angle pieces bolted together to make any angle you need. Cast some hot-dipped galvanized anchor bolts in the concrete piers or use deck supports and timbers to mount the struts. My array is mounted on a 4x6 pressure treated timber frame assembled with 1/2" all-thread. This frame provides a level base for a series of strut frames much like you have drawn. Each frame uses a vertical 10' a second 10' at 60 degrees (winter optimized) and a third cut in half for the base of the triangle and an intermediate brace. There are only 2 diagonal braces (one at each end) to prevent the frames from falling over sideways. The horizontal rails and modules stabilize the intermediate frames. The modules are attached through the slots in the horizontal strut with 1/4" stainless hardware. The modules are about 3' x 5', I think. It's been in place since 2016 with only periodic checks to make sure the bolts in the timbers are still tight. Pics taken during construction attached so you can see the various connections and orientation of the struts.

In the middle of the summer, the sun actually rises and sets behind the array since it is set so close to vertical. Eventually I'll add some vertical west facing modules to provide more power right up until the sun sets. I'm way over-paneled for summer except for late in the day when I'd like to run a small A/C until dark.

Previously, I made a rack where everything was timber except the horizontal struts for the modules. It was also on a hillside and getting all the timbers aligned was much more difficult than making a level platform and using the strut frames.
 

Attachments

  • NewPVArray_1.jpg
    NewPVArray_1.jpg
    213.5 KB · Views: 123
  • PV_Rack_Top1.jpg
    PV_Rack_Top1.jpg
    114.4 KB · Views: 125
  • PV_Rack_TimberFrame1.jpg
    PV_Rack_TimberFrame1.jpg
    214.8 KB · Views: 115
  • PV_Rack_RearBase1.jpg
    PV_Rack_RearBase1.jpg
    135.7 KB · Views: 113
  • PV_Rack_Overall1.jpg
    PV_Rack_Overall1.jpg
    163.1 KB · Views: 116
  • PV_Rack_Mid1.jpg
    PV_Rack_Mid1.jpg
    106.6 KB · Views: 111
  • PV_Rack_Full1.jpg
    PV_Rack_Full1.jpg
    187.3 KB · Views: 112
  • PV_Rack_FrontBase1.jpg
    PV_Rack_FrontBase1.jpg
    111.7 KB · Views: 115
If you can get up to a foot of snow, I would suggest putting it a bit higher above ground. This would also allow for some wind pressure to flow through the bottom vs. being directed upwards into the panels.
 
If you can get up to a foot of snow, I would suggest putting it a bit higher above ground. This would also allow for some wind pressure to flow through the bottom vs. being directed upwards into the panels.
Good call, and not just for the wind. Originally there wasn't going to be a lower row of modules but then I got greedy ;-) Now, every time it snows I 'get to' not only clear the modules but also shovel below them. There have been a couple of times when I had to shovel the piles on the left further away to prevent shading on the lower edge of the array. At least now I can get the tractor in there.

There are a few more modules on the barn roof. Unless you can reach them from the ground if you need to clear snow that's a bad idea. They are not very productive in winter since the snow can stay on them for a long time. Going up there is out of the question.
 

Attachments

  • PV_Rack_Snow1.jpg
    PV_Rack_Snow1.jpg
    202.7 KB · Views: 76
Good call, and not just for the wind. Originally there wasn't going to be a lower row of modules but then I got greedy ;-) Now, every time it snows I 'get to' not only clear the modules but also shovel below them. There have been a couple of times when I had to shovel the piles on the left further away to prevent shading on the lower edge of the array. At least now I can get the tractor in there.

There are a few more modules on the barn roof. Unless you can reach them from the ground if you need to clear snow that's a bad idea. They are not very productive in winter since the snow can stay on them for a long time. Going up there is out of the question.
I did the shoveling the first year after installing the panels and living in Ohio, even a 3-4" snowfall would give a large pile of snow to clear. I am about 2 1/2ft above the ground now and this year we've had very little snow.
 
Thanks for all the feedback everyone, this is great. Updated design, cribbing off of @OffGridICF's design. Also I'm back at 35* now that I'm not using specialty angular fittings. The winter production boost seemed negligible and I like the lower profile for wind load and accessing the top of the rack without a ladder.

Shaved $150 off the cost estimate too. :cool:


20230131-3x4_0010_jpg.jpg20230131-3x4_0020_jpg.jpg20230131-3x4_0030_jpg.jpg20230131-3x4_0040_jpg.jpg20230131-3x4_0050_jpg.jpg20230131-3x4_0060_jpg.jpg20230131-3x4_0070_jpg.jpg20230131-3x4_0080_jpg.jpg20230131-3x4_0090_jpg.jpg20230131-3x4_0100_jpg.jpg
 
Concrete piers are in. 12" dia sonatubes 2ft deep into decomposed granite (frost depth is six inches here). Two 10" jbolts in each pier to hold the bottom strut rail in place.

2023-02-14 12.06.09.jpg
I have a bit of grading to do to ensure the piers don't turn into puddles.

I'm getting the next steps sorted while waiting for it to cure. Here's a first pass diagram of the panel wiring - thoughts welcome.

20230214_panelwiringdiagram.jpg

We're running two 150v SCCs. The rack is 160ft from the battery room. Not ideal! But it's where we had to go to avoid shading and being generally in the way.

At 6awg and Voc=114v the voltage drop for 160ft is 3.7v, or 3.25%.
Vmpp is 91v, a drop of 3.5v and 3.76%.
Going up to 8awg hits 6% voltage drop. That much 4awg is... economically unattractive.

I've never looked much into voltage drop along long runs because all my previous builds have been <10ft. I found this article interesting, challenging the 2% rule.
 
Ground mounts are pretty easy to have some tilt adjustment to change for seasons to squeeze a bit more output.
 

Attachments

  • Solar Radiation Data Bakersfield CA for Flat-Plate and Concentrating Collectors.pdf
    91.8 KB · Views: 26
You only need #10 AWG grounding conductor. You do not add your circuits together for table 250.122. You would use the max current in any given circuit.
 
We're running two 150v SCCs. The rack is 160ft from the battery room. Not ideal! But it's where we had to go to avoid shading and being generally in the way.

At 6awg and Voc=114v the voltage drop for 160ft is 3.7v, or 3.25%.
Vmpp is 91v, a drop of 3.5v and 3.76%.
Going up to 8awg hits 6% voltage drop. That much 4awg is... economically unattractive.

I've never looked much into voltage drop along long runs because all my previous builds have been <10ft. I found this article interesting, challenging the 2% rule.

Hey @AxelHeyst, the piers look great. I'll ping you the next time I'm ready to expand my array ;-)

Just to clarify, if the array is 160' from the charge controller, your total wire length is 320'. You will drop the same amount of voltage on both the positive and negative conductors. Calculating the Imp as 31.65A using 2880W/91V and running the numbers in my spreadsheet I get the following for copper conductors:

Volts​
Amps​
Power (W)​
Gauge​
ohms/1k'​
ohms/ft​
feet​
ohms​
P_Loss (W)​
Vdrop​
% Loss​
91.00​
31.65​
2880.15​
12​
1.98​
0.00198​
320​
0.6336​
634.7​
20.1​
22.0​
91.00​
31.65​
2880.15​
10​
1.24​
0.00124​
320​
0.3968​
397.5​
12.6​
13.8​
91.00​
31.65​
2880.15​
8​
0.778​
0.000778​
320​
0.24896​
249.4​
7.9​
8.7​
91.00​
31.65​
2880.15​
6​
0.491​
0.000491​
320​
0.15712​
157.4​
5.0​
5.5​
91.00​
31.65​
2880.15​
4​
0.308​
0.000308​
320​
0.09856​
98.7​
3.1​
3.4​
91.00​
31.65​
2880.15​
2​
0.194​
0.000194​
320​
0.06208​
62.2​
2.0​
2.2​

Losing almost 160W from each array seems like a lot to me. 320W lost is 65% of a 480W module or $160 dollars worth of modules at $0.50/W. That said, another module might be less expensive than the larger wire. Have you priced aluminum wire? It would be OK in this application as long as you don't nick it during installation. If the aluminum conductor gets wet underground that area can quickly become an open circuit. Otherwise, just be sure to use aluminum rated terminals. I even recently learned that the current aluminum alloys do not require use of No-Alox based on the NEC. Local inspectors may still require it and it won't hurt to use it anyway.

There is also a good video from Mike Holt on array grounding. He encourages a single point ground for the system. If you have a ground rod at the array and at the inverter (or main panel) a lightning strike can create a large voltage between the ground rods which can be bad. Check this thread for the video and lots more discussion on grounding: https://diysolarforum.com/threads/conductors-in-conduit.39235/post-503520
 
I've never looked much into voltage drop along long runs because all my previous builds have been <10ft. I found this article interesting, challenging the 2% rule.

I say 2% loss at max current is OK, if wire doesn't cost too much.
But it is OK to design for 25% loss at max current. It's only money, and if you can spend it on silicon instead of copper to get more power (especially on poor sun days), you come out ahead.

Voltage drop matters between SCC and battery, battery and inverter, inverter and loads.
 
Many thanks for everyone pointing me in the right direction on this!

Here's a quick calc on 6awg vs 4awg:
pv run calc.png

Losing 315watts sounds bad, but the thing to compare it against is not no losses but the losses at a baseline design of 4awg. 'Only' losing 117watts over baseline sounds less terrible.

Then, comparing the cost premium of 4awg wire ($442) vs the amount of production I'm losing to line loss if I go 6awg ($87) seems to seal the deal for me. (I haven't found AL wire yet, though I'll look deeper).
 
Only that high a loss at high noon (or whatever your panel orientation) and optimal season and clear skies.
Look up the curves for insolation vs. season and time of day. Loss goes as current squared, and drops rapidly below max current.
Over the year, much smaller percentage loss. Convert to dollars and compare to utility rates. Compare to PV panel cost (per kWh over lifetime.)

You should include a ground wire, sized based on 1.56x Isc.
(Maybe code relates it to your oversized current carrying wire gauge; I'm just going by practical limits.)
 
You only need #10 AWG grounding conductor. You do not add your circuits together for table 250.122. You would use the max current in any given circuit.
Not sure about that; when you oversize phase on small wires you generally need to oversize the ground to match as well. The ground would technically need to be #6AWG.
 
Then, comparing the cost premium of 4awg wire ($442) vs the amount of production I'm losing to line loss if I go 6awg ($87) seems to seal the deal for me. (I haven't found AL wire yet, though I'll look deeper).

The aluminum has a higher resistance but you could go with 4awg aluminum instead 6awg copper and get a little less loss.
Copper​
Volts​
Amps​
Power (W)​
Gauge​
ohms/1k' – Cu​
ohms/ft – Cu​
feet​
ohms​
P_Loss (W)​
Vdrop​
% Loss​
91.00​
31.65​
2880.15​
6​
0.491​
0.000491​
320​
0.15712​
157.4​
5.0​
5.5​
91.00​
31.65​
2880.15​
4​
0.308​
0.000308​
320​
0.09856​
98.7​
3.1​
3.4​
91.00​
31.65​
2880.15​
2​
0.194​
0.000194​
320​
0.06208​
62.2​
2.0​
2.2​
Aluminum​
Volts​
Amps​
Power (W)​
Gauge​
ohms/1k' – Al​
ohms/ft – Al​
feet​
ohms​
P_Loss (W)​
Vdrop​
% Loss​
91.00​
31.65​
2880.15​
6​
0.648​
0.000648​
320​
0.20736​
207.7​
6.6​
7.2​
91.00​
31.65​
2880.15​
4​
0.408​
0.000408​
320​
0.13056​
130.8​
4.1​
4.5​
91.00​
31.65​
2880.15​
2​
0.256​
0.000256​
320​
0.08192​
82.1​
2.6​
2.8​

I found single conductor direct burial aluminum for $0.71/ft foot. No affiliation, haven't ever purchased from them and the shipping may be cost prohibitive:


The local Lowes shows 4awg aluminum USE-2 for $0.61/ft but they only have 77'. It's out of stock to order online but you might find some in a store near you.
 
Back
Top