diy solar

diy solar

Distance to solar array limitations

beckkl

New Member
Joined
Jun 9, 2020
Messages
232
Hello everyone,
My wife and I are building an off-grid home in the U.P, of Michigan, and the way things are looking, it appears that the solar array would need to be roughly 700feet from the house. I am going with the assumption that putting the batteries and such near the array and running AC to the house is going to be extremely expensive. I also would rather have the equipment in the house. I see they make 600V charge controllers, would that be the way to go? I assume that if I need a fairly sizable system (we plan on running AC and other normal household appliances), that I would just need to find a happy medium in configuring the array to ensure I'm not going over 600V per? I'm also assuming I would need a few of the charge controllers to meet the needs.

Anyway, I just wanted to see if the way I am thinking about this is correct?
 
Most of the high voltage DC inverters I've seen are designed for grid tie only. There is a 400v version (storedge) that can use battery however its proprietary 400v batteries so cost is $$$.

Did you have a budget?
some napkin math ...12,000w solar (with mythical 600v inverter) online wire size calcuator says 12,000w/600v -20a over 700ft with 2% loss is roughly 6awg wire. 1000' spool of 6awg is $600. So your looking at just $1200 in wire between the inverter and array. not including trenching, conduit etc.

I'd recommend you get some onsite advise from a local solar installer to give you some ideas for all the options you should consider.
 
Last edited:
Not sure what you mean, I see several options out there:


Budget this point is less of a consideration. I am assuming at this point I will have to pay an arm and a leg for cabling. I just am looking to confirm that going DC all the way is probably going to be the most cost-effective.
 
I think long run in DC is better than long run in AC, unless the usage model is PV generates during 5 effective sun hours but AC is drawn as steady load for 24 hours (lower current AC, no surge peaks.)

AC has to be in spec. DC into SCC can tolerate significant voltage drop. The guideline of 3% or 5% is just a guideline. You can design a PV array with SCC that has 30% power loss of IR drop at peak sun, so long as max Voc and min Vmp into MPPT values are met. PV panels are cheaper than copper wire, so optimize your design and purchase accordingly.

Yes, 600V charge controllers look like a good way to go. My 600V input GT inverters get about 380 Vmp, which is somewhat higher than Vpeak of 240VAC.

You don't have to go with SCC from PV to DC. If you use AC coupling like my system, PV goes to Sunny Boy GT inverters (or I think any inverter with "Rule 21" "Frequency/Watt" should work). I have moderately long DC runs, about 150' of 12 awg for 7A Imp.

Those GT inverters go to Sunny Island 48V battery inverters, which form the island grid and can connect to either generator or utility grid when available.

The other thing I've done is way oversize PV array, undersize battery (which is AGM, is kept floating most of the time). But mine isn't full-time off-grid. It is just for power failures. Larger battery or backup generator would be desirable for full-time off-grid.

Check out the DC solar trailers (2x SI 6048US + 2400W PV & Midnight Classic) that are available. If you got one with diesel generator, and if it's two 50 kWh forklift batteries are still healthy, you'll have a great starting place including turn-key power for building. Just add more PV with Sunny Boy to expand it later.
 
high voltage is were you want to be yes. AC vs DC.... I know high voltage DC switching is non-trival so 600v DC disconnect at the panels may be a challenge.
AC feeder for 700' also won't be primary voltage (240v) It would require transformers on both ends. However 600v (or higher) are fairly common and your local electrician may find them much easier to deal with. However DC->AC->transformer->transformer I cant imagine is going to be very efficient.

High voltage systems are not very DIY friendly so we just don't see a lot of the equipment/ in use here... I know i have zero experience with it.

Other considerations.... doing the inverter/batteries at array site requries a shed at that location. This could be a benfit if your going to include a backup generator in the design and dont want it sitting right next to the house. (great for noise, not great when you have to truge through the snow to figure out why the power went out)

What is site access? Difficult "trail" up a steep hill side to the array? or an easy drive across the field....
 
Sorry, I should have been more clear. We have cleared an 1.5 acres about 700 feet from where we want the house/cabin. The house is going to be in fairly dense forestry, and there will not be the opportunity to have any meaningful solar anywhere other than where we cleared the land.

When I said AC, I should have said A/C (air conditioning). My goal is get everything electric if possible, and just keep adding solar capacity as the panels are cheap now. The idea is also to have a fairly sizeable battery bank so I can use a small mini-split for both heating and cooling in two bedrooms.

Site access is fine, as we brought in heavy machinery to clear the land. But we are several miles from grid power, and likely will always be. I'm still learning all of this, but have a lot of residential electrical experience, including service upgrades and panel installations. Also have been a small electronics tinkerer for many years, so this all seems pretty doable to me. I installed a 12V solar system in an RV last year, so now I have the itch to plan out a 48V system. My father and brother are also electricians by trade so it seems pretty dumb to not DIY this when you are a DIY'er by nature ;)

Anyway, I had originally just figured I could throw gobs of panels into the field we cleared, and then would be a matter of just making sure I have them configured properly to ensure I'm not overloading the SCC, which made the 600V units look appealing. After that, aren't things fairly straight forward? Add some batteries and decent inverters and I should be off and running? I know that is oversimplifying things, but I think you get what I mean.

What I don't know is what is out there product wise that could potentially help. I see a lot of folks talk about the Sunny Boy and Sunny Island products, but I'm not sure if that is in the context of grid-tie use.
 
Sunny Island and Sunny Boy are good for both grid-tie (backup during power failures) and off-grid.
SMA is huge in off-grid systems. They also dominated grid-tied for a while, like 50% of the market while 20 other vendors split up the remaining 50%. Today, there are many competitive grid-tied inverters. There are a few other tier-1 battery inverters and many economy brands.


Working with 600 Voc DC systems is easy enough. With MC cables, no need to ever put a screwdriver on a hot terminal. Current flow is interrupted by opening a switch or breaker. Disconnect both positive and negative wires from PV array at MC connectors. Confirm with DMM there's zero volts on the wiring and hardware. (Before disconnecting MC I also use DMM to observe Voc is present, not Vmp; that way I confirm I did turn off the correct switch.)

Breakers for 600VDC are available from Midnight and Schneider. Most are two to six poles mechanically ganged and wired in series.
Heavy-duty safety switches with 3 poles, good for 600VAC or DC and 30 amps per pole are available.
Touch-safe fuse holders for 600VDC and up to 30A fuses are also available. (Don't remove fuse with current flowing, to avoid arc)

I check voltage and polarity before closing switch and connecting an array to the inverter or SCC. If paralleling multiple strings I check voltage between them first, make sure it is near zero.
 
Yikes, just looked at the prices of those. Not cheap. Look nice though. Thanks for the reply.
 
Quick question, are there advantages of AC coupling when grid-tie is not a factor, and a significant amount of your power use will be after the sun is down?
 
Quick question, are there advantages of AC coupling when grid-tie is not a factor, and a significant amount of your power use will be after the sun is down?

Consumption while sun is up (A/C, water pumping, scheduling laundry for then) provides highest efficiency, minimizes battery size requirement, combines power output of both PV an battery inverters.

Once the sun is down, neither AC nor DC coupling will differ. While sun is up, DC coupled battery charging might be 98% to 99% efficient, and AC coupled battery charging might be 97% to 98% efficient, due to each component's losses. You'll have to look up efficiency of different systems you consider.

MPPT performance varies considerably between brands and models. Some array and lighting conditions can have a power/voltage curve with local maxima, and some MPPT fail to find the highest. I don't think we have much of a "consumer's report" comparing them. Each vendor toots his own horn and contrives special cases where his product is best.

Some GT inverters have 2 or 3 MPPT, better for some conditions of multiple PV strings with partial shading.


Battery inverter efficiency varies with load percentage and input voltage (as to GT PV inverters). SMA gives curves for these, with Sunny Island peaking around 25% load. Sunny Boy have less rolloff at 100%, because that's an expecting operating condition.


One benefit of AC coupling (and some DC ecosystems like Victron, probably also some hybrid inverters) is that battery charge current can be regulated to exact target amount (assuming excess PV is available). Especially for wet-cell FLA, ideal charge current gives longest life. My system has enough PV to exceed 0.5C charging, but is programmed for 0.2C. Relays programmable according to SoC can enable or shed loads.

Typical DC system dumps onto battery all current available from PV, and inverter draws what it wants. That may be acceptable for lithium, within some maximum charge rate. Victron uses a shunt to measure battery voltage and monitoring gizmo can control SCC.
 
I can see you putting together a viable system using Schneider off-grid components. I have a XW+6848, and I'd be confident that it would meet your needs, including air-conditioning or a well-pump. I don't have the Conext 600V controller though. I used a less expensive Midnight 200. Let's assume you put together an array of 12 grid-tie panels wired in series 30V X 12 = 360V at 7-8amps. The voltage drop works out to be...
less than 10V. Not too bad. Assuming 7 amps arrives at the home at 350V, that's almost 2500W. That would work out to be ~45 charging amps at 54V. Two 1000ft rolls of 10 gauge solar cable works out to be 382$ X 2 = 764$. BTW, I tried plugging in 15 panels, but at freezing the Voc would be reaching 600V, this controller's limit.
1612214872895.png
 
Last edited:
Yikes, just looked at the prices of those. Not cheap. Look nice though. Thanks for the reply.

MSRP is quite expensive, about $5000 per 6000W Sunny Island.
Due to the DC Solar bankruptcy, many of these are being sold off new in the box, or on trailers.
I picked up mine for $1250 each (bought a pallet load), mostly they're seen on eBay around $2500 but sometimes cheaper.)

Given the room, consider a DC Solar trailer. I don't know what prices they went for at the auction, but offered on eBay and Craigslist could be $8000 and up. That includes 2x Si 6048, 1x Midnight Classic 150, 2, 50kWh forklift batteries, 2400W PV. Sometimes a diesel generator, sometimes a light tower. Those batteries, if in good shape, had been $5k msrp so $10k worth. All on a heavy trailer (about 8000 pounds or so to tow)

As a bundle, would be a good package and a bargain price.
SMA is premium quality, and if you can get over the up-front price (or get a steal), it is well worth it. They have high surge capability to start motors like well pumps. You might spend $20k for a trailer plus 10kW of additional PV panels, mounting hardware, and Sunny Boy inverters if you wanted 60 kWh/day in the summer (trailer as it comes should give 12kWh). Probably 20 years trouble-free operation is reasonable for their battery and PV inverters, although there is always the possibility of one failing along the way. I hear forklift batteries can also serve people that long.
 
I can see you putting together a viable system using Schneider off-grid components. I have a XW+6848, and I'd be confident that it would meet your needs, including air-conditioning or a well-pump. I don't have the Conext 600V controller though. I used a less expensive Midnight 200. Let's assume you put together an array of 12 grid-tie panels wired in series 30V X 12 = 360V at 7-8amps. The voltage drop works out to be...
less than 10V. Not too bad. Assuming 7 amps arrives at the home at 350V, that's almost 2500W. That would work out to be ~45 charging amps at 54V. Two 1000ft rolls of 10 gauge solar cable works out to be 382$ X 2 = 764$. BTW, I tried plugging in 15 panels, but at freezing the Voc would be reaching 600V, this controller's limit.
View attachment 35506
The problem is I don't see how I can get without running in parallel. The voltage drop jumps quite a bit at 18A. I'm looking at 6-8KW in panels to start. I may just bite the pullet and run two sets of 10AWG.
 
The problem is I don't see how I can get without running in parallel. The voltage drop jumps quite a bit at 18A. I'm looking at 6-8KW in panels to start. I may just bite the pullet and run two sets of 10AWG.

How much voltage drop at 18A, and what voltage is Vmp?
You can live with higher percentage loss than recommended, it's just watts. Voltages have to remain within MPPT and Voc limits.
Any time the sun is off angle, current and losses decrease. Losing a higher percentage for just a portion of the day (and year) can be an acceptable tradeoff.

How many PV strings? They don't have to all face the same way, even going to a single MPPT. Multiple angles reduces current and I^2R losses.

10 awg? Could pull 8 awg. I used that for one run where I paralleled two PV strings. All the rest are 12 awg for single string. As I change out panels I plan to add fuses and parallel 3, 7A strings on the 8 awg run.
Lots of 12 awg runs gives me the most flexibility. Only issue I see with it is greater derating according to NEC for many conductors in one conduit.
 
Yeah, I'm going to have to stew on it. It is in the U.P of michigan, so the winters are going to be tough. I'll look into maybe some different angles. I also would love to mount these in a way that I can adjust the angle during the winter. I've seen some folks even put them nearly at 90 to keep snow off and take advantage of reflection off the snow. Lots to consider. Perhaps I will just keep my ear to the ground on folks selling surplus wire. I was able to make out like a bandit when I built my last house by buying partial spools of romex off craigslist.
 
Here is the link to the voltage drop calculator I used. https://www.calculator.net/voltage-drop-calculator.html

Looking at 16A at 360V you get....

Not too bad with 8 gauge wire. Two 1000Ft rolls will run you about 1100$. Keep in mind though that this controller has a 80amp limit. So the maximal solar you can run through 1 controller is... 80A X 52V charging X 1.25loss factor = 5200W. You can of course add a second controller, but that's another 1100$ pop right there.
1612293617960.png
 
16A x 14V drop = 224W lost.
You can buy 250W panel from SanTan for $30 to $50. Compared to $1100 of wire ... I say use smaller gauge at lower cost, allow more loss and heating in the wires.

3% or 5% loss is arbitrary, not required for PV DC. I'd say allow 30% loss if it saves money for a given wattage output.

What does your calculator say using 12 awg? I think it'll deliver 320V.
 
You can buy 250W panel from SanTan for $30 to $50. Compared to $1100 of wire ... I say use smaller gauge at lower cost, allow more loss and heating in the wires.

What does your calculator say using 12 awg? I think it'll deliver 320V.
I would tend to agree. Adding one more panel nulifies the loss from voltage drop. The odd panel will make the array look funny though?

1612305024040.png
1612305216497.png
 
Back
Top