diy solar

diy solar

Ground mount system - extending distance

BillSF

New Member
Joined
Aug 2, 2021
Messages
33
Hi all,

I am looking at a 3 acre property with the house near the bottom of a tree-covered hill (modest grade)

I would like to clear the trees near the top of the hill, setup a ground mount system (3 to 5 kW) and run the power down to the house. I would ballpark the distance at 400 to 500 feet.

How far can I have the panels from the house before the wiring gets overly expensive? Is there some sort of relay box I can setup at a midpoint that would economically reduce the cost of heavy gauge wiring?

Any tips in general for configuring this system (higher voltage to reduce transmission losses?)
 
You would benefit from getting a high voltage solar charge controller (SCC) or one of the all-in-ones (inverter + charge controller) that can handle high voltage. Then you can put more solar panels in serial, making for higher voltage and lower current. The current is what causes loss over the wire back to your home. So you already have the idea: Trade higher voltage for lower current. That will reduce the wire costs.
 
This seems like an instance where you would need to use high voltage DC which would push the same power with less amps and lower line loss.

THe wires I use for my solar panels are rated to 600 VDC, but I have only pushed that to 70 volts. A 10 AWG wire can push 30 amps, and at 60 volts, that’s 1800 watts, but at 600 volts and 30 amps, that’s 18,000 watts. I don’t expect you’d go that high though.

A high voltage DC system is a bit out of my league. The circuit breakers I use are rated for 150 VDC or 300 VDC, but there’s a bit more to it than that.
 
There are several calculators on-line to tell you what the loss will be for a given amps on a given wire for a given distance. What you are looking for is to keep the loss relatively low. 3% both ways is a standard, but now panels are so cheap compared to wire that you may do well to put in more panels and accept a 7%-10% loss on the wire, rather than spend all your $$ on wire.
 
You can tolerate a higher percentage voltage drop in the DC wires. Doesn't have to be <3% or <5%. It's a tradeoff of copper cost vs. silicon.
By paralleling two PV strings of different orientations, peak current is lower and hours of production are longer, so reduced I^2R losses
Some panels and inverters are rated 1000V. Not meant for US residential, but you might be able to use them.
 
Couldn't you put the inverter up with the panels and then run AC to the house at 240v?

The system I'm putting in (11.7kw), uses micro inverters and is located 225' from the house. Also depends if you are going to be off grid or grid tied. - Mark
 
Couldn't you put the inverter up with the panels and then run AC to the house at 240v?

The system I'm putting in (11.7kw), uses micro inverters and is located 225' from the house. Also depends if you are going to be off grid or grid tied. - Mark

Placing the inverter at the array end would mean using a higher voltage, however it must be remembered, that the maximum demand current of the house needs to be taken into consideration and conductors sized accordingly. When a split phase inverter is used there would be a need for an additional neutral conductor to carry the unbalanced load current back to the inverter.

The better option, in my opinion, would be to use a 600V SCC' such as the offerings from Morningstar or Schneider because the maximum current would be determined by the arrays capacity, this can be used to determine the conductor size with a simple formula, for example say the array is 5000V, divide by the voltage of the series string eg. 525V the maximum current would be 9.5A, enter that into a voltage drop calculator for a 500 foot run with 10 AWG, the result is <3% standard.

Screenshot_2021-08-28-19-25-13.png

Although these controllers are in the neighborhood of $1000, the price will likely be offset by wiring costs and ballance of system if using alternate methods. Both Morningstar answer Schneider are good products, I have both, but given the choice Morningstar would win, simply due customer support, Schneider is terrible, for the end user at least.
 
Our PV array is about 250 cable-feet from the building w/ the inverter/batteries.
I did what the wonderful folks here suggested, put a bunch of Pv panels in series to get 350vdc per string.
Will be adding more this winter to each string to get about 440vdc.
Ran #10 copper in conduit, minimal voltage drop to the controller:

 
I have three separate arrays that are 500-600ft one way distance. I use Victron 250V charge controllers and bought 2000ft of 2-2-4 aluminum Stephens cable from https://nassaunationalcable.com. No need to go to 600V which complicates things (no longer can use breakers in your combiner, riskier/less tolerant connections) if you use larger gauge cable to compensate. All of this was put in a 11/4"HDPE conduit for extra protection from Chapman electric who sells that in 1000' reels. The arrays are approximately 2.5kW each but plan to add on to each one to increase them to about 7.5kW each. With the 2AWG AL I used and the current 2.5kW array (155V, 10A) going 600ft it is only about a 2.5% voltage drop. Even tripling array size to 30A will only be about 7.5% drop.

The cable was $1840 for the 2000ft and $594 + freight per 1000ft. reel of HDPE. I use a 3 pole Square D power distribution block to connect the 3 AL cables coming in. I do this since it would not be good idea to try to connect the stiffer large AL cable directly to the charge controller. I use highly flexible copper cable on other side of distribution block to the charge controller. All aluminum connections use NOALOX.
 
No need to go to 600V which complicates things (no longer can use breakers in your combiner, riskier/less tolerant connections) if you use larger gauge cable to compensate.

You can do 600V breaker with a 4p ganged breaker. Here's Midnight, and Schneider breakers show similar configurations. By having two of the ganged poles for each of positive and negative, it completely disconnects the array.


You can also use touch-safe fuses and a "heavy duty" 600VDC switch.

600V vs. 250V means less than 1/2 the current, less than 1/4 the power loss for a given wire gauge. Typical wire is rated 600V. You save on expensive copper, and no additional plastic insulation is needed so more cost effective.
 
^^
Yep, fuses in the Midnite solar and in the Schneider charge controller, and a 600vdc disconnect, no breakers.
midnite solar combinerand squared_disconnect copy.jpg
 
You can do 600V breaker with a 4p ganged breaker. Here's Midnight, and Schneider breakers show similar configurations. By having two of the ganged poles for each of positive and negative, it completely disconnects the array.
That's good to know.. I don't remember seeing that when I planned this.
600V vs. 250V means less than 1/2 the current, less than 1/4 the power loss for a given wire gauge. Typical wire is rated 600V. You save on expensive copper, and no additional plastic insulation is needed so more cost effective.
Yes, that is true but it was still more cost effective given copper prices 1-2 years ago. Today it is much worse since 2AWG aluminum is about $0.50/ft and copper 4AWG is $2.00/ft. If you are referring to the HDPE conduit as the "plastic insulation" I need that no matter if it was copper or aluminum because of the gophers, ground squirrels, etc.. even if it is over 2ft down :)

.
 
I have three separate arrays that are 500-600ft one way distance. I use Victron 250V charge controllers and bought 2000ft of 2-2-4 aluminum Stephens cable from https://nassaunationalcable.com. No need to go to 600V which complicates things (no longer can use breakers in your combiner, riskier/less tolerant connections) if you use larger gauge cable to compensate. All of this was put in a 11/4"HDPE conduit for extra protection from Chapman electric who sells that in 1000' reels. The arrays are approximately 2.5kW each but plan to add on to each one to increase them to about 7.5kW each. With the 2AWG AL I used and the current 2.5kW array (155V, 10A) going 600ft it is only about a 2.5% voltage drop. Even tripling array size to 30A will only be about 7.5% drop.

The cable was $1840 for the 2000ft and $594 + freight per 1000ft. reel of HDPE. I use a 3 pole Square D power distribution block to connect the 3 AL cables coming in. I do this since it would not be good idea to try to connect the stiffer large AL cable directly to the charge controller. I use highly flexible copper cable on other side of distribution block to the charge controller. All aluminum connections use NOALOX.

Actually I don't follow the logic that 600V complicates things, higher voltage reduces current, there likely won't be the need for a combiner, just a single breake,r which are commonly available at the voltage in question, riskier and less tolerant connections? lower current at higher voltage is easier to control, arc welding uses low voltage, high current for a reason, it's easier to sustain an arc, as such circuit breakers are less prone to failure at higher voltage.

The use of aluminum will bring the costs down when using 250V, but would be equally so at higher voltage, however 10 AWG copper would be approximately equal, or lower in cost to the aluminum required at a lower voltage and would fit in a much smaller conduit, further reducing costs.

As to the reference that 600V is high voltage, that's commonly misconstrued, 600 VDC is in fact considered low voltage, 1500V and above is considered high voltage in DC circuits.
 
As with anything technology related the longer you wait the more improvements/options you have. Even 1-2 years ago the options for going over 250V were extremely limited. I was going to go with the "600V" route originally with a Sol-Ark until the more I looked into it and decided it was bad idea for my priorities. I also waited an extremely long time for Midnite to come out with their 250V+ controller and it still isn't available. My post only was to indicate that it is possible to do it that way and given someones particular wants or concerns that is an option.

Edit: I should note that it does easily reach -45F in winter here so my Voc does require me to be quite a bit lower than a lot of people.
 
Why not just hook it up as normal and put the grid tied inverter at the house? Run the pos and neg cables from the panels at 450 volts back to the inverter 500 feet away. 10ga wire results in under 3% loss.

Keep it simple, use common components that are cheap.

DO NOT buy your wire at any big box store like Home Depot.. find an actual wire distributor.

1630166765432.png
 
One thing left out is if you have to deal with shade and you have large number of panels in series for that 450V+ voltage that shade is going to drastically reduce output of that entire string.
 
If a single series string, shading of one panel only causes loss of power from that one panel (plus diode drops)
That will be equal or higher production as compared to two parallel strings going to a single MPPT input, with one panel shaded.
 
If a single series string, shading of one panel only causes loss of power from that one panel (plus diode drops
That will be equal or higher production as compared to two parallel strings going to a single MPPT input, with one panel shaded.
Why have I been told differently so many times? :) now I look like an idiot :p Was this always the case?
 
You look like a galaxy to me. A galaxy far, far away and long, long ago.

The power/voltage curve of a PV panel ideally is one hump, so MPPT algorithm staring and Voc easily finds and tracks the peak.
Partial shading of a single cell, or partial shading of one string in parallel with another, can cause the power/voltage curve to have two humps, and MPPT algorithm might stop at the first, lower, peak it finds. Some are better than others.

There are situations of two parallel strings, one with large portion shaded, that would produce significantly less than if each string had its own MPPT. If one has say 2 panels with sun and the other has 5 panels with sun, forcing them to same voltage causes them to be well off maximum power point. (If MPPT algorithm is good enough it will find the maximum that can be produces with them in parallel.)

Some panels can't handle extend shade on one area while others have full sun, due to bypass diode overheating.

Under what conditions would a configuration like 5s2p deliver more power than 10s? (not counting efficiency vs. voltage curve of SCC.)
 
You would benefit from getting a high voltage solar charge controller (SCC) or one of the all-in-ones (inverter + charge controller) that can handle high voltage. Then you can put more solar panels in serial, making for higher voltage and lower current. The current is what causes loss over the wire back to your home. So you already have the idea: Trade higher voltage for lower current. That will reduce the wire costs.
My controller is 80A and will take up to 2000w of panels with max 145voc would 10 ga. Wire be ok if the panels are 200’ away?
 
Back
Top