diy solar

diy solar

Reverse PV Combiner Box?

Oh, my bad. Is Victron 450/200 the highest voltage they offer?
Some other brands have 600V SCC. I use 600V GT inverters. there are 1000V inverters.



You don't want 250V, even lower. At least stick with 450V and calculate Voc carefully.
I think you can use 8 panels in series not 7, but double check.

Normally I tell people to make one string face SE, another SW, and connect in parallel. They don't peak at the same time, give more hours of production, so better use of SCC and wire. But in your location probably due South is the only thing to consider.
Yeah I've got to have all my panels facing directly south, straight up and down (for the winter months due to snow and low sun angle). The whole reason for this massive and expensive system (at least to me) is to try to get as much of the four hours of sun we get on December 21st and translate that into electricity when it's cold outside. To operate our heat pump. Because I'm tired of the near death experience that comes with me cutting down trees and then chainsawing them up, things I do with the same amount of expertise as I have demonstrated in the realm of electricity.

The reason we have a 500 foot run from the PV to the cabin is because the cabin is on our one gorgeous spot (ocean & mountains & glaciers but zero sun for two months out of the year) and the one spot that sees the sun year round is 500 feet away.

So here I am. Considering paying $15,000 for 32 x 500 foot wires. In the hope that I can translate four hours of sun a day into not cutting firewood.
 
Last edited:
I think I detail most of those questions in this thread.


This thread was meant to ask just the one question about a fictional reverse combiner box. Which doesn't exist. Likely because electricity doesn't behave like I was imagining. Even so, I'm going to make one and make bazillions of dollars. I'll be sure to post the Kickstarter as soon as I figure out how electricity works and how to get solar to my cabin without paying $15,000 for 32 wires...
I see your challenge. The 22K of solar panels is a lot of power and the wire distance between the panels and battery & solar controller costs much money. :) The problem is the MPPT function per port (tracker) is 8Kw causing you to use 4 pairs of wires.

To be honest it may be cheaper to put the battery, charge controller next to the solar panels and use underground wire for 120/220 volts A/C to the remote locations. using 120 volts may help. I am guessing you only need a 100 amp service in the cabin allowing you to use smaller wire and not copper if you need to save money.
 
I see your challenge. The 22K of solar panels is a lot of power and the wire distance between the panels and battery & solar controller costs much money. :) The problem is the MPPT function per port (tracker) is 8Kw causing you to use 4 pairs of wires.

To be honest it may be cheaper to put the battery, charge controller next to the solar panels and use underground wire for 120/220 volts A/C to the remote locations. using 120 volts may help. I am guessing you only need a 100 amp service in the cabin allowing you to use smaller wire and not copper if you need to save money.
Yeah, was going down that road as you might have noticed in the other thread. We have a GSHP that needs 230V at the cabin (or will, once we buy it) and folks trying to explain to me phases and such tripped my IQ c/b. I went into that discussion thinking I knew how to size wires for DC or AC, and left having no idea how many wires I would need or what size or how to calculate it.

So my wife is going to register an account tomorrow and explore that option and step up transformers and step down transformers and we're gonna delay our order of Victron stuff. She's smarter than me and her fresh set of eyes should be good. And when we end up freezing to death after spending our life savings and getting something wrong, at least I won't be solely to blame.
 
FWIW, I see where the idea of a reverse combiner box would come from.

Looking at it like a miniature grid.
Where the solar array is the power plant.
Transform up to big voltage and send across the tall towers along relatively small and few transmission lines to a transfer station at the next town. Transformers split and distribute more usable power through the town.

Wasn't transmission the issue where Westinghouse and AC won out over Edison and DC in the Current War?
 
So here I am. Considering paying $15,000 for 32 x 500 foot wires. In the hope that I can translate four hours of sun a day into not cutting firewood.

32 x 500' = 16000'
(plus some for safety ground)
Half those wires are presently used, half for future expansion. You can lay multiple PVC conduit with rope threaded for pulling more wires later.

I think you can use 12 awg, peak current produces 5% loss. With 10 awg, lower loss.
Silicon is cheaper than copper.
2500' spool 10 awg THHN about $860 https://www.lowes.com/search?searchTerm=10+awg+2500'
12 awg about $560 https://www.lowes.com/search?searchTerm=12+awg+2500'
Should cost $4000 or $6000 in copper if you put in all 32 wires, half that if 16 wires for now.

If you wanted to get a bit cheaper, 10 awg could support two parallel PV strings connected (note 2) and somewhat higher loss, only half as many wires used.

Victron note 2) "A higher short circuit current may damage the controller if PV array is connected in reverse polarity."
If your panels are 10.23A Isc, double check polarity, and test operation with one PV string before connecting 2nd string in parallel


I suggest 12 awg. With one PVC conduit per Victron 450/200, you have 4 pairs, 8 current carrying conductors plus ground per conduit. NEC derating to 70% of the wire's 30A ampacity is good for 21A. PV panel Isc 10.23A x 1.56 factor = 16A. All good.
About $2000 in wire for now, with only the wire you need today installed. Put in several conduit for additional PV and other purposes.
 
I know this thread is about reverse combiner boxes but since that question has been answered already I'm going to address the bigger picture.

The defining problem I see is distance.
q: Is there also a problem of terrain?
a:
What I mean is, imagine having to cover those 500 feet when the power goes out in the middle of the night in the worst weather imaginable.

q: Is this setup long term enough that you will have to cover that 500 feet in the worst conditions when you are old and infirm?
a:

I suggest you and/or you wife prepare a scale map that shows relative position of all the relevant buildings, the pv array and the heat sink.
Doesn't have to be fancy, it would also be useful to know how difficult it is to trench on each of the possible wire runs depending on soil composition and topology.

Please don't call the locations A, B, C... it just adds to the cognitive load.
Descriptive names make it easier for us to efficiently think and communicate.
For example, lets name the PV array... "PV array".
Other examples might be "main house", "rental cottage" and "power/utility building".

This is a large scale project in a remote harsh environment so its important to get it right.

And I will ask one last time...

q: Have you done a power audit?
a:
 
Last edited:
32 x 500' = 16000'
(plus some for safety ground)
Half those wires are presently used, half for future expansion. You can lay multiple PVC conduit with rope threaded for pulling more wires later.

I think you can use 12 awg, peak current produces 5% loss. With 10 awg, lower loss.
When we do this online calculator for DC wire runs, we were using the no more than 3% voltage drop. It says 3%-10% voltage drop is not great but okay for lights and non-sensitive loads. If we go the DC route 500' to the SCCs, I guess what would be important for us is knowing if the 450/200 SCCs are non-sensitive loads. I'm guessing they are and that between the SCC and the MultiPlus the power that would be delivered to the batteries, GSHP, and cabin electrical panel would be refined appropriately by the Victron equipment. So if that's the case, I guess we can accept higher voltage drop in exchange for using smaller wires. Something we need to think about, so thanks for sharing that.
 

Attachments

  • drop.jpg
    drop.jpg
    20.7 KB · Views: 2
q: Is there also a problem of terrain?
What I mean is, imagine having to cover those 500 feet when the power goes out in the middle of the night in the worst weather imaginable.

q: Is this setup long term enough that you will have to cover that 500 feet in the worst conditions when you are old and infirm?

q: Have you done a power audit?
It's a nice level walk from our cabin to the PV and when I trench I'll be making a nice path as I do it.

This setup is definitely long term. This is our home. We need the power so we'll have to do what we have to do. We'll have a separate generator at the cabin if we need to use it.

We are in the process of a power audit. We have a general idea.
 
I'm glad you didn't lay out a stack of cash prematurely.

Yeah me too. Was a good idea posting this thread just to have all you mean people with your fancy knowledge and "been there done that" breaking me rather than me breaking the bank. That was just the kind of response I was hoping for. Once the wife's account gets approved by an administrator (MrsAlaskanNoob is the handle) she'll try her hand at understanding the original scenario (SCC, batteries, inverter co-located with the PV array and AC power sent 500' to the cabin & potentially using step up/down transformers) since that was too much for me to comprehend.

Then we'll argue about what's the best option.

Then she'll win.
 
Last edited:
Have you run pvwatts for your location? I suspect if you download the hourly spreadsheet it produces you will see that the winter watts are not anywhere close to the specification for the panels. Since the watts are low, and the volts relatively higher ( cold) then the amps will be much less than Impp in the spec. This means your winter DC transmission loses will be less than you might calculate otherwise. The SCC will operate just fine with the slight loss of volts in the DC transmission and amps are unaffected by distance.
 
Have you run pvwatts for your location? I suspect if you download the hourly spreadsheet it produces you will see that the winter watts are not anywhere close to the specification for the panels. Since the watts are low, and the volts relatively higher ( cold) then the amps will be much less than Impp in the spec. This means your winter DC transmission loses will be less than you might calculate otherwise. The SCC will operate just fine with the slight loss of volts in the DC transmission and amps are unaffected by distance.
We did run PVWatts for our location. And our PV spot has no tree or terrain shading. It's the only spot on our property that is like this, it's on a south facing slope where there is a south/north running canyon so the trees to the south don't block the sun. It gets the solid four hours of unshaded sun that we get on 21 December.

For PVWatts we used premium panels but there was no selection for bifacial panels like we're using. And our panels will be perfectly facing south and will be straight up and down to optimize winter sun and to keep snow off them. On a ground mount elevated above ground snow. And I'm going to berm up some dirt to the north of the panels so that in winter it creates a snowy reflector that I'll try to optimize to reflect perfectly on the back side of the panels.

I made a spreadsheet that uses the PVWatts expected solar data for a 22KW array. I then used soil temps ten feet down for our location and average temps for our location and plotted out the expected GSHP energy usage for our cabin as it's insulated now, and for a more insulated cabin.

Based on PVWatts, we should be good to go with power although since these are averages, there will likely be days or maybe weeks where we turn off the GSHP and heat with the wood stove just so we can use the solar power for other things since it won't be enough to power the GSHP and our other cabin loads. But outside of that, we should be good to go.

More to your point, and just based on Will's test of a bifacial panel on a hot day, I'm expecting these panels are going to provide at least 100% power on 21 December as they will be non-shaded, optimized for the sun angle, and surrounded by snow (I'm going to berm up dirt to the north of the PV so the snow will fall and make a backside reflector) to reflect on both sides of the panels. But I could be wrong for sure, so that's why I want to size the wires to allow for doubling the PV if we need to.

BTW this graph is cut off but summer production according to PVWatts will get up to 96KW a day in summer. I'm hoping to have a dump load that pumps heat back into my horizontal loop ground during the summer to help re-charge the soil my GSHP will be drawing from.
 

Attachments

  • GSHP_Consumption_Production.jpg
    GSHP_Consumption_Production.jpg
    161.5 KB · Views: 1
Last edited:
Yeah, was going down that road as you might have noticed in the other thread. We have a GSHP that needs 230V at the cabin (or will, once we buy it) and folks trying to explain to me phases and such tripped my IQ c/b. I went into that discussion thinking I knew how to size wires for DC or AC, and left having no idea how many wires I would need or what size or how to calculate it.

So my wife is going to register an account tomorrow and explore that option and step up transformers and step down transformers and we're gonna delay our order of Victron stuff. She's smarter than me and her fresh set of eyes should be good. And when we end up freezing to death after spending our life savings and getting something wrong, at least I won't be solely to blame.
Good luck and keep us updated with your adventure.
 
When we do this online calculator for DC wire runs, we were using the no more than 3% voltage drop. It says 3%-10% voltage drop is not great but okay for lights and non-sensitive loads. If we go the DC route 500' to the SCCs, I guess what would be important for us is knowing if the 450/200 SCCs are non-sensitive loads. I'm guessing they are and that between the SCC and the MultiPlus the power that would be delivered to the batteries, GSHP, and cabin electrical panel would be refined appropriately by the Victron equipment. So if that's the case, I guess we can accept higher voltage drop in exchange for using smaller wires. Something we need to think about, so thanks for sharing that.

3% is a good guideline for AC wiring, maybe DC to loads.

PV feed to SCC or GT inverter is the most non-sensitive load imaginable. So long as Vmp minus wire IR drop doesn't drop below minimum MPPT voltage, everything works fine. It doesn't matter what percentage of power is lost (during times of peak power) other than that power loss itself. Even if 25% of 50% loss, things still work. For instance, you might only approach full output from your array in summer when light doesn't pass through as much atmosphere. So as array approaches 22kW, loss reaches 5% or 25% or whatever you designed for. But it's summer, you get a lot of hours, in addition to higher output. In winter, what does an insolation calculator say? If 0.7x the power, around half as much loss in wires.
 
3% is a good guideline for AC wiring, maybe DC to loads.

PV feed to SCC or GT inverter is the most non-sensitive load imaginable. So long as Vmp minus wire IR drop doesn't drop below minimum MPPT voltage, everything works fine. It doesn't matter what percentage of power is lost (during times of peak power) other than that power loss itself. Even if 25% of 50% loss, things still work. For instance, you might only approach full output from your array in summer when light doesn't pass through as much atmosphere. So as array approaches 22kW, loss reaches 5% or 25% or whatever you designed for. But it's summer, you get a lot of hours, in addition to higher output. In winter, what does an insolation calculator say? If 0.7x the power, around half as much loss in wires.
That's good to know, thank you.
 
I see your challenge. The 22K of solar panels is a lot of power and the wire distance between the panels and battery & solar controller costs much money. :) The problem is the MPPT function per port (tracker) is 8Kw causing you to use 4 pairs of wires.

To be honest it may be cheaper to put the battery, charge controller next to the solar panels and use underground wire for 120/220 volts A/C to the remote locations. using 120 volts may help. I am guessing you only need a 100 amp service in the cabin allowing you to use smaller wire and not copper if you need to save money.
I think that if he puts the battery, inverter, SCC down by the panels, it would only work with a highly insulated heated shed where he is. I am in the same boat as AlaskanNoob but not as drastic. MY AIO has 2 mppt, but the 250' run is already trenched in with 2 conductor 600v direct bury. I guess sizing my array for 1 mppt for now is all I can do, until next year when I can add panels, and trench in more wire. I was hoping to split my 4 awg 2 conductor to run both mppt with more panels.

I'm building my own battery/inverter shed, and have propane roughed into it's location. I'm installing a wall heater that will operate with and without power for a fan.

I hope it all comes together for you AlaskanNoob.

If your looking to save money on wire, contact industrial electrical companies up there. Sometimes they have excess cable, and sell it instead of storing it for another job, then forgetting about it. They might have single wire for conduit. I bought 350' 4 awg 2 conductor direct bury for $300. Not bad around here.
 
Alrighty then,
It's Saturday Night and We're LIVE.
DId you mean to post this in the HUMOR forum?
Hey Why Not, Wytreasure, if your going to respond to a thread, "Why Not" make it constructive?? AlaskanNoob is trying to solve a problem. Maybe you are but not everyone is an electrical genius.
 
Back
Top