diy solar

diy solar

Design Review Requested

pyearick

New Member
Joined
Mar 15, 2022
Messages
5
I'd appreciate feedback on the following design.

I have a number (~150) of 10 year old, ground mount Solyndra SL-200-200 panels that we want to use for grid-tied PV. These sit on feet that clip onto the panels and support them about 12 inches off the ground. I am working in a large sunlit area to place these panels. However, there is approximately 225' from this panel area to the electrical service for the home. I plan on controlling the wiring costs by putting two GoodWe inverters near the panels. The inverters will be mounted by the solar panel field and their outputs will be connected through a 125 A panel (Aggregation Panel on the SLD). Then I want to bring AC over the 220' feet back to the house through a direct burial cable. The AC will connect to an AC disconnect near the meter and connect to the service in the house either through the box or a line side tap after we see if either the main panel or the sub-panel in the basement has the capacity.

I'm concerned about whether what I have in mind work, for example, will I have too large a voltage drop given the long AC run - and anything else that I probably didn't think of!

Thanks!!

Basics
Location - Philadelphia, Ground Mount
Usage - ~26000 kWh/yr (AC, large house, barn)
Goal - Grid-Tied for bill reduction

Panels (please see the Solyndra documents attached) -
(128) Solyndra SL-200-200 These panels are approximately ~10 years old
Voc: 124.6 Voc (on the panels I've tested I'm seeing 115 V)
Isc: 2.35 A
Pmp: 200 W
Vmp: 91.7 V
Imp: 2.18 A
Max System Voltage: 1000 V
Max Series Fuse: 24.4 A

Panel Array Sizing
I plan on combining the panels 4x4 (16 panels) - four panels per string and 4 strings per MPPT. I will use a junction box to combine the strings. Using all 8 MPPTs on the two inverters lets me use 128 of the best panels out the the 150 that I have.

Inverters - (2) GoodWe GW9600A-MS
Max Output Power: 19,200W
Input Voltage Range: 80-600V
GoodWe has worked with me on the string sizing - their estimates are in the GoodWe image.


Combiner Boxes for the strings (Solar Combiner Box PV Combiner Box 4 String With15A Rated Current Fuse Solar DC Breaker Lightning Arreste and Solar Connector for Solar Power Systems10AWG Solar Cable)


Aggregation Panel (SIEMENS W0816ML1125CU 125 amp, 8 Space, 16 Circuit, Outdoor Center)


AC Disconnect -

Wiring -
Panels to Junction box - #10
Junction box to inverter - #10 THHN in 3/4" PVC
Inverters to aggregation panel - #8 THHN in 3/4" EMT
220' from Aggregation Panel to AC Disconnect 2-2-2-4 Direct Burial Cable
AC Disconnect to Line Side Tap (or box after closer inspection of the box)
 

Attachments

  • 2022-05-20 08.41.28.jpg
    2022-05-20 08.41.28.jpg
    1.6 MB · Views: 8
  • 2023-09-01_10-33-01_SunPath.jpg
    2023-09-01_10-33-01_SunPath.jpg
    414.4 KB · Views: 7
  • GW_A-MS_Datasheet-EN.pdf
    507.8 KB · Views: 5
  • pdfslide.net_designguide-200series-solyndra.pdf
    3.9 MB · Views: 2
  • SiteDiagramR2.jpg
    SiteDiagramR2.jpg
    638 KB · Views: 7
  • SLD.jpg
    SLD.jpg
    602.6 KB · Views: 8
  • Solyndra-SL-200-200-2.jpg
    Solyndra-SL-200-200-2.jpg
    22.9 KB · Views: 10
  • GoodWeA-MS9600_calcs.png
    GoodWeA-MS9600_calcs.png
    87.6 KB · Views: 14
In dealing with inverters, voltage drop causes voltage rise at the inverter and too much voltage rise can cause the inverters trip on high voltage. If possible you will want to keep the voltage rise under 2% to prevent issues, for this 220' run you will need to up size the feeders to 1/0 copper or 3/0 aluminum. You can use any voltage drop calculator to do your own calculations, by enter the max output current of your inverters as a load, noting the drop is a rise in voltage at the inverter.

A couple of other issues I see:
  • PV combiner box doesn't seem to be listed, NEC requires listed components, consider Midnite solar combiners.
  • No visible listing mark on solar panels, verify listing of panels(UL, ETL, etc), NEC requires PV panels to be listed.
  • Design guide for solar mounts, doesn't provide data for a ground mount install, which may make getting permits more difficult, and possibly require review by a licensed engineer.
  • No way to install the CTs included with the inverter, may not be needed. But it may be helpful to determine if a remote meter connected via rs485 is supported by the inverter.
  • Optional, provide 1-2 cables shielded CAT6 cable between the house and the inverters, for Ethernet or rs485 monitoring or metering
 
Last edited:
If the inverters can do 600v, Wouldn't you save $$$s on wiring if you did high voltage DC from combiner box to inverters with the inverters at the house and the combiner boxes under the panels?
 
If the inverters can do 600v, Wouldn't you save $$$s on wiring if you did high voltage DC from combiner box to inverters with the inverters at the house and the combiner boxes under the panels?

I agree. The four panel strings are about 370 VMP, and with a 20kW (effective - 80% of 128 panels) energy input it'll be 54A on the long wire.

If converted to 240VAC before the long wire you'll be pushing 80A, with 50% more voltage drop, or 50% more copper at the same voltage drop.

Further, while the inverters are rated for outdoor mounting and environment, they will last longer and perform better in a better environment.
 
Looking at the design guide for those panels, see figure 6 on page 7, you don't need any combiner boxes if you can wire them in that way. The low current of these panels, means don't get any where close to the ampacity of wiring, even with 4 in parallel, so you don't need any over current protection, and the inverter provides the disconnect, so you can just run the 2 wires from these 4x4 blocks to inverter.
 
I agree. The four panel strings are about 370 VMP, and with a 20kW (effective - 80% of 128 panels) energy input it'll be 54A on the long wire.

If converted to 240VAC before the long wire you'll be pushing 80A, with 50% more voltage drop, or 50% more copper at the same voltage drop.

Further, while the inverters are rated for outdoor mounting and environment, they will last longer and perform better in a better environment.
Rough math shows 60% or about $800 cost increase for 16x10awg + ground for running all the pv circuits back, vs just bringing a single ac line on 3/0 aluminum.
 
Rough math shows 60% or about $800 cost increase for 16x10awg + ground for running all the pv circuits back, vs just bringing a single ac line on 3/0 aluminum.

Each MPPT controller has a maximum of 12.5A input, so you definitely don't need 10AWG. Maximum current should be 9.4A per pair (short circuit panel current * 4). But you are correct - I did not take into account the fact that each combined string needs its own pair of wires, and that may result in a higher cost.
 
Thank you all! I'm sorry I haven't had a chance to reply yet to all the comments. However like pvgirl said, I would have to bring back eight pairs of wires from the 8 sets of 16 panels.
 
Looking at the design guide for those panels, see figure 6 on page 7, you don't need any combiner boxes if you can wire them in that way. The low current of these panels, means don't get any where close to the ampacity of wiring, even with 4 in parallel, so you don't need any over current protection, and the inverter provides the disconnect, so you can just run the 2 wires from these 4x4 blocks to inverter.
Thanks pvgirl, that will save some dollars!
 
If the inverters can do 600v, Wouldn't you save $$$s on wiring if you did high voltage DC from combiner box to inverters with the inverters at the house and the combiner boxes under the panels?
That is exactly how I had it originally and was counting on the high DC voltage to save in the wiring cost. But it got expensive as well as the the cost of the conduit to run 220'. So I'm going to try and do it with that direct burial cable as long as I can keep the voltage drop reasonable.
 
Each MPPT controller has a maximum of 12.5A input, so you definitely don't need 10AWG. Maximum current should be 9.4A per pair (short circuit panel current * 4). But you are correct - I did not take into account the fact that each combined string needs its own pair of wires, and that may result in a higher cost.

NEC Calculates the ampacity for the wiring of each input as 14.9 amps (max Isc * 4 parallel strings * 1.56). When all 16 current warring conductors are run together we must derate the conductors, 50% per Table 310.15(B)(3)(a) requiring a conductor with an amapacity of 29.8. Which is 10 awg or 90c rated 12 awg.
 
Last edited:
In dealing with inverters, voltage drop causes voltage rise at the inverter and too much voltage rise can cause the inverters trip on high voltage. If possible you will want to keep the voltage rise under 2% to prevent issues, for this 220' run you will need to up size the feeders to 1/0 copper or 3/0 aluminum. You can use any voltage drop calculator to do your own calculations, by enter the max output current of your inverters as a load, noting the drop is a rise in voltage at the inverter.

A couple of other issues I see:
  • PV combiner box doesn't seem to be listed, NEC requires listed components, consider Midnite solar combiners.
  • No visible listing mark on solar panels, verify listing of panels(UL, ETL, etc), NEC requires PV panels to be listed.
  • Design guide for solar mounts, doesn't provide data for a ground mount install, which may make getting permits more difficult, and possibly require review by a licensed engineer.
  • No way to install the CTs included with the inverter, may not be needed. But it may be helpful to determine if a remote meter connected via rs485 is supported by the inverter.
  • Optional, provide 1-2 cables shielded CAT6 cable between the house and the inverters, for Ethernet or rs485 monitoring or metering
pvgirl:

I am definitely going to run underground capable Cat 6 to the inverters, something like this:


Do you think I need to include that cabling in the SLD for the power company?

Also, I am not sure what you were referring to: "CT"?

Thanks a lot!
 
CT means Current Transformer. That is the standard way of reading AC current, whether directly attached to an inverter or remotely via a meter that is then connected via a digital link to the inverter

(there’s probably situations where CT is used colloquially when the sensing technology is not a CT)

Note if you go with shielded CAT you should understand the correct grounding so as to avoid ground loops etc. I don’t know what the best practice is, it’s probably to ground at one end. Following to see if there’s a canonical way

Is direct bury typically 90C rated? (IE is it designed to have 90C rating when wet)
 
That is exactly how I had it originally and was counting on the high DC voltage to save in the wiring cost. But it got expensive as well as the the cost of the conduit to run 220'. So I'm going to try and do it with that direct burial cable as long as I can keep the voltage drop reasonable.
I thought there was multiconductor cable listed for direct burial for PV. Hopefully rated to 600V
 
CT is current transformer, the install manual for those inverters shows installation of ct's as part of the install, if you have any requirements from your utility provider to limit production you will require these, but in simple setups they may not be needed.

I would recommend a shielded cat6, and best practice is to ground the shield at one end, but for simplicity and just for Ethernet an unshielded cable will work. I would not think you need to include these communication cables in your single line diagram.
 
Back
Top