I'd like to set up solar for my critical needs

I'm not sure if this is a thing but I'd like to put up some solar panels and a battery system to power a portion of my on-grid home.  I don't want to do "grid-tied" for a variety of reasons.

I'm thinking if I get a half a dozen solar panels, I could tie them to a battery system and then run wires around my basement to a few key spots (well, freezer, fridge, furnace, etc).  Then if the power goes out for a week we'll still keep our food frozen and have running water.

I'd also run an outlet to the garage to charge the missus' prius whenever we have excess electricity.

My challenge is lots of trees and only one semi-decent place to mount panels.  The south side of my barn (not the roof) just above the milky white colored doors (see attached pic).  I'd have to cut down a few trees to unshade it.

I'm thinking I could fit 6 panels on the barn (maybe more if I get creative) and have them quite vertical so they shed snow and are optimized for winter sun angles up here (45 degrees N).  I don't know much about standard panels but I think that would generate something around 2000W?

The barn is 200' from the house but I have an empty 3/4" conduit between the two.  It gets down to -30F here so I suspect batteries would need to live in the house.  

I suspect I would need a charge controller, inverter and pile o' batteries to pull this off.  The AC wiring is well within my capabilities but I don't know much about DC or solar.  

Where/how should I locate these components?  Any suggestions for styles/types of components?  

I'm assuming the panels go on the barn, the charge controller is next and goes either in the barn or in the basement.  Then wires go from the controller to both the batteries and the inverter.  Maybe more wires go from the batteries to the inverter?  Then I run normal AC stuff from the inverter to my new off-grid outlets.  If the system is undersized, I can plug the equipment back into the grid while I upgrade.

While money isn't falling off trees around here, I'd rather spend a bit to get reliable stuff that I don't have to mess with.  I'm guessing that means lithium batteries instead of refurbished forklift ones.

It's cloudy here a lot in Nov/Dec so I'm happy if I can provide my minimum electrical needs (well, fridge, freezer) at that time and have extra juice in the summer for other deliberate uses (cloths washer, prius charging) as the opportunity presents itself.  So lots of battery storage may not be needed but some capacity would be great.

I don't have electrical consumption numbers for the furnace (hard wired) or the vertical freezer that we open once a day.  I do for the fridge and it's around 2kwh per day.  The shallow well pump, if my math is correct, uses 0.021 kwh per 5 gallon cycle.  I suspect we use less than 50 gallons per day so 0.25 kwh per day for the well.  If the power is out and we're home, we'd be heating with wood (no furnace needed).  If we're away and the power was out we wouldn't be running the well or opening the fridge/freezer.

So I feel like having 4kwh per day of capacity would handle us in the winter.

Thoughts/suggestions?  Thanks!

You're right about keeping the batteries warmer, they lose potential capacity in colder temps. Typical panels are around 300w+ these days, so if you got 6 you could get up to 1800w but especially in winter it'll be well below that. Each is probably 30+v and in cold weather they produce a bit more voltage, and passing clouds can have an edge effect that also bumps up voltage back and forth. So if you connect them all as a single string you'd start at 180v, and I've read you should add 20% for each of those 2 possible situations, so 252v total. That is good for being able to use a smaller wire to get that back to the charge controller, but then the controller needs to be able to handle that potential max voltage.

(the following is from memory as I'm not at home) When I built my power board, I included a breaker/fuse between the panels and the controller that was a little higher voltage than what the panels should produce, then that went into the CC. From the CC the positive went into another fuse rated for just above the rated output amps, and the batteries and inverter also connect to that on a bus bar and the batteries positive also has its own fuse; the negatives all connect to a bus bar and the ground wires have their own bar that goes outside to a copper rod in the ground, and the panels also have a ground wire to a copper rod in the ground in case a nearby lightning strike energizes anything.

My setup is also separate from the grid, although I rarely see power outages I wanted to build and learn before I was living off grid, using used batteries. When determining capacity you need to decide how many days of power you want available, then either double that for flooded lead acid storage, or increase it by 20% for lithium. Also remember that if something uses say 1kwh/day AC, there will be a conversion loss through the inverter, maybe 10-15% depending on the inverter. So now that 1kwh AC could be 1.15kwh DC. Also any time you are using AC power, the inverter has to be running and it causes a little draw too. I'm not sure how much that is.

Thanks Mark!  Is there a good resource I should be looking at that can guide me through all this stuff?  It seems like there's lots of info out there but I'm not sure what would be best for my situation.

Also, if I wanted to add panels in the future, does that mean I'd have to replace most of the other components?  Or can I size some parts to be able to handle more panels in the future?

just my two cents here.

Our setup is 6-340w solar panels. I believe they are wired in series so 3 panels are in series and than those are connected together in parallel.
The solar panels are about 100 feet from where the batteries are located. We are using Tek Cable which has 3 wires which are 10AWG.

We have 4 - 6v 370AH Lead acid batteries which are wired in series so that we have 24VDC. So 24vdc x 370ah = 8800Wh. or 8.8KWH(i think?) With them being Lead acid we only actually have 4.4kwh

We are currently running a 5.1 CF danby AC freezer and it is kicking our butt. Almost every day we need to run the generator for about 2 hours. At least right now when the sun is so low.

So needless to say i think your system wouldn't cut it to be honest Mike. Unless you are wanting to run a generator. Also that fridge is sure sucking up a lot of juice!

I think our freezer is 197KWH/Year  so 197/365 = 539wh a day.


The big thing to figure out is what will be your usage when it is cold and when there is no sun out. Also when the sun will be low.

Here is one thing which will help you out.  Energy audit and sizing tool

Another thing to look at is what voltage you need to have the minimum line loss. Voltage drop calculator

Feel free to ask other questions mike!

Thanks Jordan!  It's a full size, store bought fridge (not particularly efficient).   Thanks for the links, I'll delve into them.  By still being grid connected, I could always charge the batteries if needed until I get the system sized and working.  Assuming I can get it working.

4-340w panels run in series would be about 160VOC running at about 10AMP. So 200' of line running at 160v with 10 amp would have a line loss of about 2.97% Which i believe is acceptable.

Now if we were to add two more panel strings of 4 - 340w we would get 160voc at 30amps. Which would bring the line loss of 10AWG down to almost 9%, which is to much i believe.

If we bring the VOC up to 240v running at 20amps and using 8AWG the voltage drops to 2.60% with a 200' run

So having the batteries in your house is possible.

You can also increase the voltage even higher than 240v. There are charge controllers which will accept up to 600V

if we just run all of the 12 - 340 panels in series. We get 480VOC which running at 10amps would give a nice 0.99% voltage drop. Using 10AWG. Looks like you could even get away with 14AWG Copper

Mike Haasl wrote:Thanks Jordan!  It's a full size, store bought fridge (not particularly efficient).   Thanks for the links, I'll delve into them.  By still being grid connected, I could always charge the batteries if needed until I get the system sized and working.  Assuming I can get it working.

We turn our fridge off around october. We are lucky in that the temps stay around fridge temps most of the winter. Right now we have -7C. Most of the winter is is around 4*C outside where our fridge is.

Do you have somewhere outside where you can keep your fridge? Or an unheated part of the house?

The fridge is in the kitchen and pretty integral to our current "normal seeming American kitchen" set-up so I doubt we can do anything about that.  The freezer is in the basement in a room that's about 55F in the winter.

When I get home I can look up some links I've saved along the way. I'm seriously considering not having a fridge in the future house, but just outside the door in the root cellar, so that in the winter when the sun is low it won't need any power. Obviously not the typical American setup. You'll definitely need a larger setup if you keep a full size fridge and freezer and other bits active, or consider a more efficient setup to reduce the need for the generator running. I'm not sure if all brands are similar, but I've seen a couple inverters that allow you to add another when you want to grow so you can scale that way.

The first and foremost thing you have to do is to do an itemized audit of what you want to power BEFORE you start buying anything  The single most prevalent place where people fail is that start buying inadequate stuff BEFORE they know what they need.

Things like the frig, freezer, lights, and TV are easy.  Just running those things at my cabin I'm consuming about 2-3 kWh of power each day.  The item that you mention that catches my eye though is the well-pump.  That potentially might be the single biggest power user on your whole place.  Is it doable, sure.  I"ve run my well-pump exclusively on solar for years now.

First, is your well-pump AC powered?  Is it 120VAC or 240VAC?  Do you know the running and starting wattage?  There might be a placard displaying all this information, or you might have to determine it yourself.  I myself had to take the latter route.  I bought a "clamp meter" that has the capability to read "inrush current".  I now have two, a very expensive Fluke 274, and a cheaper Uni-T 216C.  Both can read inrush, and both are accurate to within 1% of each other.  With the inrush meter I found that while my pump needs 2230W to run, it needs about 8930W for 1-2 seconds to start.

How is your well-pump controlled?  Does it come on any time of the day and fill a bladder tank, or does it run continuously when you turn on a tap.  Does it fill an elevated tank and then shut off?

I myself have an elevated tank which gets filled during the day when I have solar, and then flows downhill at night when the power is off.  Before installing the solar system I used a generator to power the pump.

How your pump provides your water will be critical to plan a solar system that can power your pump.  The least expensive strategy might be to build the smaller system, but power the pump via generator only, but that will depend on how the pump is controlled.

Let's put together two suggested systems based on whether or not you need to run your pump.

24V System 1 can provide 4-8 kWh of power per day

6 250W grid-tie panels, wired in 3S2P (cheap on Craigslist right now.  Buy locally instead of mail order; it will be far cheaper)  75$ X 6 = 450$
4 6V Trojan L-16 lead-acid batteries  350$ X 4 = 1400$
60A Epever Tracer 6415AN charge controller  (I have the higher quality Midnight 200V controller)  270$  Midnight is ~670$
24V Sine-wave inverter  Samlex 24V 2000W for about 650$  Schneider 24V Conext 4024 4000W for1500$.  The Samlex is a cheaper high-frequency inverter with little inrush capability.  The Conext is a low-frequency split-phase 120VAC/240VAC sine-wave inverter with a built in generator charging circuit.  I have the Conext for my workshop and that is what I would recommend.

48V System 2 can provide 8-16kWh off power per day with 12 panels, or 14-23kWh with 18 panels.

12-18  250W grid-tie panels, wired in 3S2P (cheap on Craigslist right now.  Buy locally instead of mail order; it will be far cheaper)  75$ X 12 = 900$
8 6V Trojan L-16 lead-acid batteries  350$ X 8 = 2800$; 18 panels = 1350$
60A Epever Tracer 6415AN charge controller 270$;  100A Epever Tracer 6415AN charge controller 570$  (I have the higher quality Midnight 200V controller) Midnight is ~670$
48V Sine-wave inverter Schneider XW-Pro 6848 sine-wave inverter. The XW is a low-frequency split-phase 120VAC/240VAC sine-wave inverter that can surge to 12000W for 60 seconds.  It has a built in generator charging circuit.

What you might want to do besides placing panels on your roof is to build ground mounts like I have.  These single pole mounts can be rotated left to right to catch more early morning and late afternoon sun.  By rotating east to west, I can run my 240VAC well pump from 8am to 4pm with zero battery drain.

The last thing will be breakers, power boxes, and switches, that can add up to several hundred dollars.  I have a power center with all the breakers to control the both the DC and AC side of things.  

Instead of powering just a few circuits, you can also get a transfer switch or either/or breaker panel to wire the solar system directly into you house panel.  Then your power becomes transparent.  There MUST be a control that prevents your power from ever going into the grid during a blackout.  Lineman's lives are threatened if you energize a circuit they think is dead.  It might be best to have a professional electrician do this for you, because lives are at stake here.

Thanks for those excellent options Michael!  

To elaborate on the pump a bit, it's a shallow sand point well.  It's on 120V and I have a note that it's 1080W (I'm not at home at the moment so I can't look at the nameplate right now).  It fills a 5 gallon pressure tank and kicks on and off automatically based on the pressure in the tank.  When it kicks on it runs for 70 seconds, pumps 5 gallons and uses 0.021kwh (per an experiment where I watched the meter spin as the pump ran).

I don't have any elevation to work with so I'd have to use some sort of timer and a large pressure tank to allow it to not run at night.

When I get home I'll get more pump nameplate info and plug the Kill-a-watt into the freezer to check its consumption.

jordan barton wrote:just my two cents here.

Our setup is 6-340w solar panels. I believe they are wired in series so 3 panels are in series and than those are connected together in parallel.
The solar panels are about 100 feet from where the batteries are located. We are using Tek Cable which has 3 wires which are 10AWG.

We have 4 - 6v 370AH Lead acid batteries which are wired in series so that we have 24VDC. So 24vdc x 370ah = 8800Wh. or 8.8KWH(i think?) With them being Lead acid we only actually have 4.4kwh

We are currently running a 5.1 CF danby AC freezer and it is kicking our butt. Almost every day we need to run the generator for about 2 hours. At least right now when the sun is so low.

So needless to say i think your system wouldn't cut it to be honest Mike. Unless you are wanting to run a generator. Also that fridge is sure sucking up a lot of juice!

I think our freezer is 197KWH/Year  so 197/365 = 539wh a day.


The big thing to figure out is what will be your usage when it is cold and when there is no sun out. Also when the sun will be low.

Here is one thing which will help you out.  Energy audit and sizing tool

Another thing to look at is what voltage you need to have the minimum line loss. Voltage drop calculator

Feel free to ask other questions mike!

Okay  I am going to say you are expecting too much.  At best a panel is good for 4 to 6 hours a day.  Lets say you are at the 4 meaning divide by 6 for the day.  that means at best you are looking at 340 watts average for the day.  For 4 hours you are on the panels but the other 5 sets you are on battery.  If getting the power in and out of the battery costs you 20% to 30%.  Call it 20% just for the heck of it.  Now you are down to 272 watts.  Now to run the inverter it is probably costing you 20% also now you are down to 212 watts.  And the freezer will have some losses too.  Most chest freezers are down in the 1/4 hp rang so that is 190 watts.  So if your panels produced power every day and did nothing but run the freezer you would be borderline.  Now add in partial and full gray days and you shouldn't even be close.

It gets down to -30F here so I suspect batteries would need to live in the house.

Since your primary need is not in the house I would consider putting the batteries in a small temp controlled insulated enclosure in the barn near the panels. It will use some electricity to keep them warm in winter but without doing any math yet I'd be willing to estimate it will be less loss than running long wires (AC or DC) into the house.  

Mike Barkley wrote:

It gets down to -30F here so I suspect batteries would need to live in the house.

Since your primary need is not in the house I would consider putting the batteries in a small temp controlled insulated enclosure in the barn near the panels. It will use some electricity to keep them warm in winter but without doing any math yet I'd be willing to estimate it will be less loss than running long wires (AC or DC) into the house.  

All the loads listed are in the house? You'd need to put the inverter/s in the barn too, and then run AC wires to the house. I see no advantage at all, myself...

Mike Haasl wrote:Thanks for those excellent options Michael!  

To elaborate on the pump a bit, it's a shallow sand point well.  It's on 120V and I have a note that it's 1080W (I'm not at home at the moment so I can't look at the nameplate right now).  It fills a 5 gallon pressure tank and kicks on and off automatically based on the pressure in the tank.  When it kicks on it runs for 70 seconds, pumps 5 gallons and uses 0.021kwh (per an experiment where I watched the meter spin as the pump ran).

I don't have any elevation to work with so I'd have to use some sort of timer and a large pressure tank to allow it to not run at night.

When I get home I'll get more pump nameplate info and plug the Kill-a-watt into the freezer to check its consumption.

I am curious, is it on such a small pressure tank because the well only holds a small amount of water? Or some other reason?

Something seen somewhat often in off-grid setups is a large pressure tank or bank of tanks, allowing a bunch pump runtime done in daylight to carry through at least one night if not longer. Less wear on the pump, too. Ought to be a wee bit less energy used to run pump for 700 seconds than 10x70 seconds... works nice with a genset, too, in a pinch..

Michael Qulek wrote:The first and foremost thing you have to do is to do an itemized audit of what you want to power BEFORE you start buying anything  The single most prevalent place where people fail is that start buying inadequate stuff BEFORE they know what they need.

Things like the frig, freezer, lights, and TV are easy.  Just running those things at my cabin I'm consuming about 2-3 kWh of power each day.  The item that you mention that catches my eye though is the well-pump.  That potentially might be the single biggest power user on your whole place.  Is it doable, sure.  I"ve run my well-pump exclusively on solar for years now.

First, is your well-pump AC powered?  Is it 120VAC or 240VAC?  Do you know the running and starting wattage?  There might be a placard displaying all this information, or you might have to determine it yourself.  I myself had to take the latter route.  I bought a "clamp meter" that has the capability to read "inrush current".  I now have two, a very expensive Fluke 274, and a cheaper Uni-T 216C.  Both can read inrush, and both are accurate to within 1% of each other.  With the inrush meter I found that while my pump needs 2230W to run, it needs about 8930W for 1-2 seconds to start.

How is your well-pump controlled?  Does it come on any time of the day and fill a bladder tank, or does it run continuously when you turn on a tap.  Does it fill an elevated tank and then shut off?

I myself have an elevated tank which gets filled during the day when I have solar, and then flows downhill at night when the power is off.  Before installing the solar system I used a generator to power the pump.

How your pump provides your water will be critical to plan a solar system that can power your pump.  The least expensive strategy might be to build the smaller system, but power the pump via generator only, but that will depend on how the pump is controlled.

Let's put together two suggested systems based on whether or not you need to run your pump.

24V System 1 can provide 4-8 kWh of power per day

6 250W grid-tie panels, wired in 3S2P (cheap on Craigslist right now.  Buy locally instead of mail order; it will be far cheaper)  75$ X 6 = 450$
4 6V Trojan L-16 lead-acid batteries  350$ X 4 = 1400$
60A Epever Tracer 6415AN charge controller  (I have the higher quality Midnight 200V controller)  270$  Midnight is ~670$
24V Sine-wave inverter  Samlex 24V 2000W for about 650$  Schneider 24V Conext 4024 4000W for1500$.  The Samlex is a cheaper high-frequency inverter with little inrush capability.  The Conext is a low-frequency split-phase 120VAC/240VAC sine-wave inverter with a built in generator charging circuit.  I have the Conext for my workshop and that is what I would recommend.

48V System 2 can provide 8-16kWh off power per day with 12 panels, or 14-23kWh with 18 panels.

12-18  250W grid-tie panels, wired in 3S2P (cheap on Craigslist right now.  Buy locally instead of mail order; it will be far cheaper)  75$ X 12 = 900$
8 6V Trojan L-16 lead-acid batteries  350$ X 8 = 2800$; 18 panels = 1350$
60A Epever Tracer 6415AN charge controller 270$;  100A Epever Tracer 6415AN charge controller 570$  (I have the higher quality Midnight 200V controller) Midnight is ~670$
48V Sine-wave inverter Schneider XW-Pro 6848 sine-wave inverter. The XW is a low-frequency split-phase 120VAC/240VAC sine-wave inverter that can surge to 12000W for 60 seconds.  It has a built in generator charging circuit.

What you might want to do besides placing panels on your roof is to build ground mounts like I have.  These single pole mounts can be rotated left to right to catch more early morning and late afternoon sun.  By rotating east to west, I can run my 240VAC well pump from 8am to 4pm with zero battery drain.

The last thing will be breakers, power boxes, and switches, that can add up to several hundred dollars.  I have a power center with all the breakers to control the both the DC and AC side of things.  

Instead of powering just a few circuits, you can also get a transfer switch or either/or breaker panel to wire the solar system directly into you house panel.  Then your power becomes transparent.  There MUST be a control that prevents your power from ever going into the grid during a blackout.  Lineman's lives are threatened if you energize a circuit they think is dead.  It might be best to have a professional electrician do this for you, because lives are at stake here.

Good post.

An important upside of either movable arrays or multiple differently angled arrays, is that a day with 2h of late or early sun can give a LOT more juice if a few panels are aimed right. For max annual watts, the sweet spot is south, but my experience off grid has been that a couple extra KWh on an otherwise cloudy day is more useful than 10KWh on a fully sunny day.


Multiple arrays are generally best served by multiple charge controllers. Epever is very affordable; I 'd rather run something higher quality, IMO it's usually cheaper in the long run.


I think the pump is no problem compared to the fridge and freezer, myself. These will need a hefty battery bank to weather dark days, and a hefty solar array to get that bank charged quickly when the winter sun appears.


I like the idea of dedicated circuits; I would do it with an inverter that can pass through AC power and switch to battery if there js a power outage. A Victron Quattro has 2 AC inouts and as I understand it should allow for a config where everything on the connected circuits runs on solar until the batteries are down to say 80%, then it will switch to grid power. If grid goes down it will switch to battery power, and it can then turn on a generator and charge the batteries if batteries fall to yy%.

They are not the easiest to set up, and Victron expects your dealer to be your tech support, so if you go with Victron choose your dealer with care.



Michael, how long have you had your Schneider gear? Have you had to deal with their support people at all? have heard some shitty stuff about them for both support and longevity, but nearly bought one anyhow, the split phase 240v from a single unit was very tempting... I went with Victron, not 100% perfect from a software or support perspective but solid.

I have had no issues running a 3/4HP jet pump for extended periods while irrigating, off 120V from a single 3000VA Multiplus..

What are you willing to spend?

What is most important to accomplish? Is the primary goal disaster preparedness, convenience for smaller outages, or..?

A LiFePO4 bank is great. Except when it's not... my $6600 400Ah/24V (8.8KWh, ~6KWh usable in ,y conservative config) bank of Sinopoly cells is toast, after 2.5 years, for no known reason.. I have 7x 300w panels feeding it; until the battery went, it was a good system, and had no issue with my Unique brand mini-fridge and assorted modest power tool loads, even in a PNW winter. A full size fridge and freezer would need a lot more battery and panels, in my location.

(Don't buy unique brand products, quality does not justify the price in my experience!)



It is simple to add more solar, by adding more charge controllers. Some brands can synchronize, but this is not critical, especially with LiFePO4


Inverter expansion... Some brands can sync multiple inverters; that's how my twin Victron units are configured, to give 240V. I could add more matching units... But, what if the specific model is discontinued... best to buy a big one or matched set up front, IMO.


Batteries are a huge cost. Fridges and freezers can be a huge load..

If you want a cheaper option, you could consider sizing to just handle the freezer and all other loads, and plan on using frozen bottles to keep the fridge, or a better insulated cooler, down to temp. Less plug and play, but the do-all system is going to be quite a few thousands of dollars, certainly a 5-digit number unless you are ok running a genset pretty often when the grid is down.

There is definitely a reason that off-grid systems of moderate cost tend to have small fridges, and cheap ones have none!



I am a big believer in a redundant set of gear to power minor loads. I can limp along for a long time with enough power to handle charging cellphone, tablet, flashlight/headlamps, and occasionally a power tool battery.. especially in winter where there is free coldness just outside. All these needs can be handled in strictly DC form.. A daily genset run can handle the freezer and water, even if the inverter fails... don't forget to wire in a way that allows an easy reconfigure to bypass the inverter!

Mike Haasl wrote:By still being grid connected, I could always charge the batteries if needed until I get the system sized and working.

This is an idea that I've been toying with. Set up a complete solar system with batteries and a new panelboard. Then, one of the inputs to the charger is from my grid tied panelboard. This, in theory would allow me to slowly add new batteries and solar panels while moving more and more loads over to the new panelboard as I learn. Start small, grow as finances and knowledge allow.

D Nikolls wrote:All the loads listed are in the house? You'd need to put the inverter/s in the barn too, and then run AC wires to the house. I see no advantage at all, myself...

Yes, all the loads are in the house.  The panels would be on the barn.  The smallish conduit runs between the two.  So I need to transfer the power 150-200' from the solar panels to the loads.  I've been assuming the batteries should live in the basement.  Heating an insulated battery closet in the uninsulated barn would probably take a lot of electricity.

D Nikolls wrote:
An important upside of either movable arrays or multiple differently angled arrays, is that a day with 2h of late or early sun can give a LOT more juice if a few panels are aimed right. For max annual watts, the sweet spot is south, but my experience off grid has been that a couple extra KWh on an otherwise cloudy day is more useful than 10KWh on a fully sunny day.

Multiple arrays are generally best served by multiple charge controllers. Epever is very affordable; I 'd rather run something higher quality, IMO it's usually cheaper in the long run.

With my equipment, I have not found that to be true.  One controller handles 6 arrays just fine with a Midnight combiner box.  Each individual string of panels gets it's own breaker so all are electrically protected.

I am what is called "over paneled".  That means my a solar arrays can produce more power than the controller is designed to handle.  I keep some arrays pointed SE, while others are pointed SW.  That keeps the noontime amps within the amp limit of my Midnight controller.  As a fail-safe though, my controller is programmed to clip off at 65A for "just in case".  The extra panels help out on the cloudy days that would shut down other systems.  I can still produce ~600W in the rain, and I've never even come close to depleting my battery bank.

D Nikolls wrote:
Michael, how long have you had your Schneider gear? Have you had to deal with their support people at all? have heard some shitty stuff about them for both support and longevity, but nearly bought one anyhow, the split phase 240v from a single unit was very tempting... I went with Victron, not 100% perfect from a software or support perspective but solid.

I have had no issues running a 3/4HP jet pump for extended periods while irrigating, off 120V from a single 3000VA Multiplus..

The answer is 5 years and two years.  I installed my 48V cabin system in 2016.  That's with a XW+6848 Schneider inverter that puts out split-phase 120/240V.  I liked Schneider so much that I installed a second Conext 4024 in my workshop.  All my equipment as worked flawlessly for this time, so I've never had to call tech support.  Yes though, I hear it's not very good.  It seems that Schneider is focused primarily with large professional installations, and they don't want to be bothered by DIYers

One tool every off-gridder should have is a clamp meter that can read "inrush current".  The one I like now is the Uni-T 216C.  When set to "inrush", it can measure the starting current that a motor-driven appliance draws in the first half-second after turning on.  Your pump is 120V?  Mine is 240V.  The refrigerator is FAR easier to start, with a startup surge of about 500W.  The well-pump has a startup surge of ~8900W.  These are the values I've measured.  Without measuring it, I'd guestimate your pump is drawing 3000-4000W right at startup, then drops down to about 1000W running.

This clamp meter is a bit pricy at 75$, but it has features that cheaper meters don't provide.  A budget meter can only read AC current.  The higher priced ones read both AC and DC current.  Next in the cost scale is inrush, and up at the top is reading RMS current.  I don't think RMS is something an off-gridder is ever going to need though.

Ouch Mike. I'm going to have to declare brain flatulence for the insulated enclosure thought. I totally misread your post last night. I thought it said you weren't going to use the solar for the house except during power outages & would charge the Prius at other times. See what watching two 3 year olds all week will do? Brain flatulence indeed.

Michael Qulek wrote:

D Nikolls wrote:
An important upside of either movable arrays or multiple differently angled arrays, is that a day with 2h of late or early sun can give a LOT more juice if a few panels are aimed right. For max annual watts, the sweet spot is south, but my experience off grid has been that a couple extra KWh on an otherwise cloudy day is more useful than 10KWh on a fully sunny day.

Multiple arrays are generally best served by multiple charge controllers. Epever is very affordable; I 'd rather run something higher quality, IMO it's usually cheaper in the long run.

With my equipment, I have not found that to be true.  One controller handles 6 arrays just fine with a Midnight combiner box.  Each individual string of panels gets it's own breaker so all are electrically protected.

I am what is called "over paneled".  That means my a solar arrays can produce more power than the controller is designed to handle.  I keep some arrays pointed SE, while others are pointed SW.  That keeps the noontime amps within the amp limit of my Midnight controller.  As a fail-safe though, my controller is programmed to clip off at 65A for "just in case".  The extra panels help out on the cloudy days that would shut down other systems.  I can still produce ~600W in the rain, and I've never even come close to depleting my battery bank.

D Nikolls wrote:
Michael, how long have you had your Schneider gear? Have you had to deal with their support people at all? have heard some shitty stuff about them for both support and longevity, but nearly bought one anyhow, the split phase 240v from a single unit was very tempting... I went with Victron, not 100% perfect from a software or support perspective but solid.

I have had no issues running a 3/4HP jet pump for extended periods while irrigating, off 120V from a single 3000VA Multiplus..

The answer is 5 years and two years.  I installed my 48V cabin system in 2016.  That's with a XW+6848 Schneider inverter that puts out split-phase 120/240V.  I liked Schneider so much that I installed a second Conext 4024 in my workshop.  All my equipment as worked flawlessly for this time, so I've never had to call tech support.  Yes though, I hear it's not very good.  It seems that Schneider is focused primarily with large professional installations, and they don't want to be bothered by DIYers

One tool every off-gridder should have is a clamp meter that can read "inrush current".  The one I like now is the Uni-T 216C.  When set to "inrush", it can measure the starting current that a motor-driven appliance draws in the first half-second after turning on.  Your pump is 120V?  Mine is 240V.  The refrigerator is FAR easier to start, with a startup surge of about 500W.  The well-pump has a startup surge of ~8900W.  These are the values I've measured.  Without measuring it, I'd guestimate your pump is drawing 3000-4000W right at startup, then drops down to about 1000W running.

This clamp meter is a bit pricy at 75$, but it has features that cheaper meters don't provide.  A budget meter can only read AC current.  The higher priced ones read both AC and DC current.  Next in the cost scale is inrush, and up at the top is reading RMS current.  I don't think RMS is something an off-gridder is ever going to need though.

The theory is that with different orientations feeding one array, there will(may?) be multiple 'peaks' visible to the controller, and it could lock on to the wrong one, meaning reduced power.

I have also read that shifting sun/shade conditions are more challenging for the controller with differently oriented arrays..

So... in practice does it matter? Don't know for sure!

There isn't a ton of discussion that I have found; there are people stating that they do it, with a quality controller like midnite or victron, and it works fine, but I haven't seen examples where they actually compare the total power generated between a single vs multiple controller config with the same panels. So... would anyone know if they were losing a modest percent of their potential power, without a comparable available?


Completely agree that if you are going to do it, overpanelling is a great fit! (Victron says this is OK too, but definitely something to check before doing it on a given brand/controller..)



I don't like single points of failure, so I will run multiple controllers anyhow. Although, one could make the case for running a single controller with a matcher safely sealed away in a box...


I'm glad your units are holding up well, 5 years is long enough to count for something.


I have the exact same Uni-T meter, but must admit I haven't checked the inrush current on my temporary pump. The Victron units are rated for 6KW startup current... once I finish running the buried 240V lines I will switch to a larger 240V pump, using both Victron units for 12KW max surge rating.

D Nikolls wrote:
The theory is that with different orientations feeding one array, there will(may?) be multiple 'peaks' visible to the controller, and it could lock on to the wrong one, meaning reduced power.

I have also read that shifting sun/shade conditions are more challenging for the controller with differently oriented arrays..

So... in practice does it matter? Don't know for sure!

What I can tell you is that I've been doing this way for 5 years now, and I've never observed the controller getting confused.  I think though that the single most important thing is that all the arrays are matching voltage-wise.  One important point though is that this is with Tier-1 Midnight controllers.  I don't know if a economy controller like Epever is sophisticated enough to do the same.

Whats interesting though is that I've managed to incorporate rather different panels that work well together even though they output different voltages.  With my single row panels, I started out with 72-cell grid-tie panels putting out ~38.5V each, so three in series makes ~115.5VDC.

I've gotten great deals more recently with 60-cell grid-tie panels putting out ~29.5V.  I built a two-row array facing due West that has 4 245W grid-ties.  Four panels in series makes 118VDC.  Meeting together at my combiner, the two voltages are so close that the controller doesn't notice.  The important thing to remember though is that these are just numbers on a sheet of paper.  In the real-world you see surprises.  I can remember one morning when I was checking voltages string by string, and I found the three panel strings putting out 124V, and the four panel strings putting out 110V.  So, in the end, the controller doesn't even seem to notice, and is happy to accept all the raw power given it.

I got a bit more well pump info.  It's a 3/4 horse pump.  If my measurements in the first post don't match what the motor plate says, I'd love to be alerted to my mistake!

That's great Mike.  With that information we could go with final system design.  Anticipating a 4X surge, I think the Conext series of inverters will handle your pump.  Here are their specs.

Going with your statement above about the 4kWh of useage per day, we can build a system that handles that.  Let's say you want two days of automony before the batteries drain to 50%.  The math is....
4000Wh X 2 Days X 2fold battery capacity = 16000Wh of power.  To store that much power, you'd need either a 667Ah 24V battery, or a 333Ah 48V battery.  That is just about the size of my 48V Trojan L-16 battery bank.  L-16s are one of the standard off-grid batteries, so they are easy to locate at most battery shops.  If they don't have them in stock, they can quickly order them.  I paid about 350$ each for mine.  Call that 2800$ for eight of them.  Alternatively, you could go to Costco and buy their 6V golf-cart batteries at 99$ each.  You'd need 16 of them wired in 8S2P but they'd only cost 1600$ plus core-charges.

You'd need about 2800W of solar to keep the L-16s happy, and somewhat more 3100W to the two strings of GC batteries.  That would be two arrays like I pictured above, with 1400-1600W of panels per array.  What I am seeing on Craigslist right now in my area is 250W panels selling for ~75$, so 3000W would be ~900$

To handle 3000W/50v charging = 60A so I'd get a 80A charge controller.  The Epever 8420AN costs 360$ right now.
https://www.ebay.com/itm/165132856429?_trkparms=ispr%3D1&hash=item2672af6c6d:g:VJAAAOSwB3phHfgB&amdata=enc%3AAQAGAAACoPYe5NmHp%252B2JMhMi7yxGiTJkPrKr5t53CooMSQt2orsSLY2M1Gjmuwt9c03vWNfiRhIPQcAGEIe4mgiVQvbRmmTmrScScsmS%252BheqxFRjQ1zJj0l9bR%252BFFD97fXnLM7VkPfj4ejM%252B75CpZ39zC7qg3DCQXT%252F7GkkARs2rqHllg9vY%252FqusilQctofG0N%252Fo6LKGgpRJMiA0wkrUCFlQKasS4O%252B5rg5jP9wfi%252Fx7aJAUWcgEAbNkB0vv5UuSrLzb1yPUFSpn%252BhHGO6A1968UAUmonWMQ9twMqTC4NukiL6x27HSiIzOpJXcaw%252F4KDASJpxTAcT9D%252F%252BsZNu6uwya8Lk5wHFQ1cUnPmOIqfm6ojlq79%252FTJ2Xnx0UrRE96goCVRkKUfXgoBUP12rk0HJVi%252BgjY0HiYHFSCL0T2ZxCmRV6reDeUq%252BY7D8X5PtrLFymMTrmIbNsizSzOLXKszm0gj9iMSsx6ARlgL68LBshg8IJpEpDuiwZX%252BHA6OHbn6Y1oNSeuvGZ5PFWuZzHKzlv2krqsY41AzpULRpx7%252FksM9rVZ%252B6byl6zWuk8NBCVC3DcHUj4q3Xt9HvN5uZgtbMgiVpJT28WRpLoC41wAWGKCq%252FrkjHTLREcMxw8nq83%252B89Jzu3YoZP7SALhiOn4%252FwuRVI6C6lt%252FDNjYtXJAq6YInL0QlNFflicKOLOG1KCWQS2p8ATJuAiZx3aqNXMGiHf0%252FVqSrRr513yzH5Y4Xqo5aKP8chpVD%252FS47xpAnDkx1ITNyBH5mxyPFLO5oPxL%252BbqmFmCWRHmfI5v2v%252FvN2oSglmcFiyGcT82Bh8hTRn3OwPGJyEkNUbfVyUwcE8Ey4q3lXMNbOF7j0Hw%252FUaKxVczIb8Jt5Cr4DW5DZJ%252F2irT6ZhYTMCfCvw7Q%253D%253D%7Cclp%3A2334524%7Ctkp%3ABFBMkM60ucNf

Finally, the inverter, a Conext SW 4048 which is 1600$ right now.
https://ressupply.com/inverters/schneider-electric-conext-sw4048-120240-invertercharger
Throw in a couple of extra hundred dollars for breakers, wire, and connections.  With your panels wired 3S4P, you will need a combiner box with four breakers to keep up to code.  I have this one.
https://ressupply.com/electrical-distribution-parts/midnite-solar-mnpv6-combiner-box
Make very sure that you only use DC breakers for your solar strings.  AC breakers would be dangerous because they can NOT handle DC arcing.

Thanks Michael!  I have the kill-a-watt plugged into the freezer right now, I should have some usage info for that this evening so I can verify that load.

Would you put the charge controller in the basement (150' from the panels) or in the barn?

Mike Haasl wrote:Would you put the charge controller in the basement (150' from the panels) or in the barn?

I would put the charge controller closest to the batteries. You want to minimize the length of wires for the part of the system with the lowest voltage.

Thanks Jeff!

Just checked the freezer and it's used 0.32 kw in 27 hours.  So 0.284 kw per day

Mike Haasl wrote:Thanks Jeff!

Just checked the freezer and it's used 0.32 kw in 27 hours.  So 0.284 kw per day

If you are installing 2800-3000W of panels to keep the batteries happy, then even in December with about 2.5 sunhours of light you should be able to make 7000Wh of power (7kWh), so at that point your freezer will be almost insignificant.  Just 4% of your total power.

Awesome, so to summarize, I'd need 12 panels ($900), $360 charge controller, $1600 inverter, $2800 batteries, $300 misc, $200 more for wire I forgot to buy, which puts me around $6200.  Plus mounting the panels but I'd cobble that with stuff I have laying around.

The biggest problem is finding room for 12 panels.  I think they're around 26" by 60", right?  

Anybody familiar with the work of http://livingenergyfarm.org/? I am doing some remote works for peasant farmers in Western Asia (biointensive micro-farming with syntropic add-ons inside an ecosystem restoration project) and housing/energy are big issues. I am attracted to the zero-inverter, nickel-iron batteries, daylight drive approach from LEF but have no experience to base this attraction on. Anybody out there who has a connection and an opinion? Thanks.

Subscribing to this thread as im in Rural Wisconsin as well.

In addition to finding a good book, or three, before you start, definitely spend a lot of time watching this guy's YouTube channel: https://www.youtube.com/c/WillProwse

He has tons of good info, from breaking down system designs to in-depth reviews of different brands of products.  A wealth of knowledge.

You can never spend too much time in R&D before you start committing your money, labor, and expectations to actual equipment.  That has been my experience.  The last thing you want to be doing is "learning on the job" when you've just bought $10K in equipment.  Having said that, you will no doubt still learn a bunch in the process.  But a major design modification mid stream is bound to end up costing you coin and heartache.

My own system is grid-tied, and I've not much experience with it.  So I will refrain from offering too much specific advice.  But two things come to mind...

First, I see a lot of posts where you are calculating watt-hours.  That is of course good and necessary to figure out your total daily, monthly, yearly usage vs total generating capacity.  That way you will know if you can produce enough power for your needs.  But that is only half of it.  Be sure to also keep a close eye on your instantaneous power draw in watts.  There is no point in having a system that will supply your annual power needs, as averaged out on paper, if in practice it won't actually run your equipment.

For example, I currently use an electric tankless instant water heater.  It is necessary based on the tiny shack in which I currently live, which could not accommodate a large water tank.  In winter when I have it maxed out, it draws 9 kW.  That's a huge amount of power!  Now, I don't run it very often - it doesn't feed my dishwasher or clothes washer, and how many minutes per day am I actually running hot water in my sink or shower?  Not too many.  And of course the whole point of a tankless heater is that it only consumes power when the water is flowing.  So, the total contribution in kWh to my total annual usage is not so huge.

But that hardly matters if the combination of batteries and inverter I use will only output 7 kW continuous power, then I can't run my 9 kW water heater, period.  Yes, I said I am currently grid-tied.  But I plan in the future to likely acquire batteries and go off-grid, so I have already started to research the equipment and its limitations.  When/if I do go off-grid in the future with this same inverter and the batteries that are currently compatible with it, I know I will have to switch to a regular tank heater (not a problem, as by then I'll be in a larger building).  It may well use more watt-hours in total, but my system capacity can likely provide for that.  The key is that it will draw only a fraction of the watts at any one time compared to the tankless.

If you have no single, large-draw appliance on your system, then it is only a matter of adding up the peak power usage of your individual appliances.  If your freezer starts itself up - and remember that the power draw is massively larger on startup, even if only for a matter of seconds - and your car is plugged in, and your well pump is starting itself up too, will the total combined wattage be provided by your system?  If not, then 1) consider different equipment for your system; 2) look into different appliances that draw less power; or probably easier to 3) use timers on your appliances to ensure that the limited combination running at any one time won't max out your system.

Second, have you considered the possibility of saving a shit-ton of money on your system by omitting the batteries altogether?  You would likely still run your power through a single, smaller battery just as a buffer to even out the fluctuations in output, but this could still be a lot cheaper than having a big battery bank that provides real storage.

Obvious pro: a lot of $ saved.  Equally obvious con: you could only power your equipment while the sun is shining.  But since you aren't trying to run your whole house off of this system, but rather only a few specific appliances, it might work for you.

Consider that chest freezers are very good at staying cold.  It is totally okay to run your chest freezer for just a couple of hours each day (if even that much), and leave it "unplugged" for the next 24 hours or even longer.  I've done it: 3-day power outage and my chest freezer wasn't even close to thawing.  And depending on your daily schedule, plugging in the electric car during the sunshine hours might be just as feasible as plugging it in overnight.  At least on some days of the week.  Consider those days a bonus, and on the remaining days plug it into the grid power from your house.  You'd still come out way ahead on a dollar basis.

Now, I've never lived with well water, so I don't know if only running a well pump intermittently makes sense or not.  If it fills a large enough pressure tank/holding tank while it is running, perhaps you could go a day or more before it needs to run again...?  If you currently have a smaller holding tank, would adding a larger tank be prohibitively expensive?  I just have no idea.  If you were considering this strategy, it would help to adopt a habit of shifting the timing of some "optional" water-intensive household activities.  People need to shower when they need to shower.  But refraining from running the dishwasher or the clothes washer until the middle of the day, when you know the solar system is powering the well pump, wouldn't be that difficult.

Good to remember that if your solar system is providing real-time power in this way - as opposed to primarily just recharging a battery bank - it is all the more important to size your panel array around the minimum winter daily output.  So, you might end up with a slightly larger array.  But that's okay, since solar panels are still way cheaper than batteries.

I'm subscribing to this thread too.  We're in New York State at about the same latitude.  So many cloudy days that there's a term "Ithacization" used to describe the condition of not seeing the sun for weeks at a time.  Our land was off-grid when we bought it but (long story) the power company ran lines up our street so we are now grid-tied.  We'd like to have solar back-up since in the year since power reached us, there's been frequent outages.

My husband is the one investigating solar.  Seems like the priciest part is the batteries.  He's doing further research on DIYing the battery bank.  Can't be super-helpful on that one yet but it may be possible and much more cost efficient.  Anyone know anything about that?  We'll need enough to power the well pump and the bare essentials for at least several days with minimum recharging if the weather is bad.  Considering some wind powered charging since we have fairly steady wind but yet to tackle looking into that.  We're guessing it could be tied into the solar setup but could be wrong.

This guy's youtube channel seems very helpful:

https://www.youtube.com/watch?v=atYZ4RtUJhU

Last March we installed a grid tied system on our roof.
It covers about 1/3 of the south side of our house.  We are roughly 37° 42' N
This past summer our house was noticeably cooler and more comfortable.
So, I would encourage people to put the panels on the roof if they live in an area where the snow wouldn't build up there.
The electric bill we paid for this last December was $11.53.  That was all just taxes and fees.  We haven't paid for any actual electricity since August.  If we didn't air condition, we wouldn't have paid for any electricity since April.
We don't have battery backup.  But, if you can pull that off with your budget, I would encourage you to go for it.

Hi Andrew.  I have 2 x 20amp hour nickel iron battery setups from Living energy lights.  They work well.  I put a 58 farad super capacitor on one and it will run a small 100w inverter.  I have talked by email with Alexis and I like the way he does things.  He uses direct power for quite a few things on their farm.  He has a great video showing his shop tools running on daylight.

Now, I don't actually speak electricity--my husband is the one who figured out our system and helped several others in our area (WV, not too different than you) set up others. Ours is off grid but he helped others do grid tied and one like yours with separate systems. I have learned a few things:
I agree with those who say Figure out what your needs are and how to minimize them before sizing your system. We were going to get one of those thousand dollar small fridges intended for offgrid systems, but we used a Kill-a-Watt to check out our stuff and found that the fridge we had bought for temporary use--a used one for $200, fairly new in 2009-- only uses about 1 kwh/day, a little less in winter and more in summer.
We started with T-105 golf cart-type batteries, expected to last 5 to 7 years. In the seventh year we got some expensive, heavy, 2 volt lead acid tubular batteries, made by Discover Energy, that the specs claim will be down to 80% efficiency in 25 years. So they should last the rest of our lives. But $4000 for 12 of them--worth it but not cheap. We have six panels, 1540 watts.
Most important, in case you haven't considered it, depending on  your place, the best place for your panels isn't necessarily a roof. Ours are mounted on pressure treated wooden posts 60 feet from the house--the house is at the edge of tall trees for summer shade, the panels are out in the sun all day. They're easily adjusted seasonally, and can easily have snow brushed off in winter, something I would NOT want to do while hanging out a window when it's ten degrees.. Nor would I want to not do it--you need MUCH more power in winter, or rather you need what you do get much more because days are shorter and it's much cloudier. But if there is snow on the ground, it does reflect significantly more light onto the panels--you don't want to waste that!  Brushing off panels while standing on the ground is much preferable. The mounting system my husband designed is detailed on his website at spectrumz.com, Going Solar. It's been copied by lots of people. By the way we have a well pump--my husband tries to always run it when the sun is shining, which saves the energy lost when power goes into the batteries and then back out. Our water is stored in four 55-gallong plastic drums upstairs and runs by gravity feed down to a kitchen and bathroom sink..pressure tanks take quite a lot of power I think.
One more thing--because we are on the ridge and have that aforementioned shade, we don't need AC--on hot days we most use tiny fans my husband made out of pair of old computer fans. Just a few watts, and it's a small stream of air but it's enough.

Phil Swindler wrote:Last March we installed a grid tied system on our roof...
The electric bill we paid for this last December was $11.53.  That was all just taxes and fees.  We haven't paid for any actual electricity since August.  If we didn't air condition, we wouldn't have paid for any electricity since April.
We don't have battery backup.  But, if you can pull that off with your budget, I would encourage you to go for it.

Yes, if you are looking for a whole-house system - very different than what the OP is proposing - then it is often very attractive to go grid-tied.  Glad that your own experience has been so good!  Batteries are great for self-reliance, but unless your power needs are very, very low, buying a suitably large battery bank can be cost-prohibitive.  Even in the best circumstances, batteries will certainly be the most expensive part of your system.  If I were to buy lithium batteries for my own small (3.8kW), whole-house system right now, enough to ensure off-grid capacity for my needs, they would cost at least twice as much as all other system hardware combined.

However, a word of warning to any newbies who might be reading this thread: your grid-tied mileage may vary!  If you're interested in grid-tied photovoltaics, then before even wasting your time researching other aspects of your system, look up what net-metering rate schedule is offered by your utility.  Without good net-metering terms, you may decide to go another way.  And those terms vary enormously from utility to utility, based on what your state government has forced them to offer.  I say "forced," because the general rule is that your utility really doesn't want you to generate power, even while you sell them the excess for cheap; they are invested in the traditional model of selling you electricity that they distribute from a centralized generation facility.

With growing federal requirements to transition to renewable energy, you'd think they might be happy for you to add grid-tied solar - for which they get to claim the federal renewable generation credits - to their system at your own expense, right?  Nope.  They are building they own solar farms.  So they intend to remain within the traditional model: they generate power centrally, albeit renewably, they control the distribution, you buy their power.

My own utility is switching from a very attractive net-metering schedule, to an almost-as-nice interim option, to just this month a new option going forward that is so unattractive it is practically punitive to the home generator.  How is that for boldly going into a renewable energy future?!  With a little luck, I should be grandfathered into the interim schedule for years to come.  But if I were to end up on the newer "punitive" rate schedule, it could be so bad that I literally would save little-to-no money even while generating most of my own power!  Yes, that bad.

The point is that net-metering can make or brake your plans for a grid-tied PV system.  If I didn't already have a solar system, and my only choice was the newer rate schedule, I might well consider investing that money somewhere else entirely.

Full disclosure: my own economic situation relative to utility bills is somewhat unique simply because I consume so little energy.  A more typical American household that consumes at least two, or maybe three, or even four times more energy than I do would likely still find at least some benefit to home generation, even under my "punitive" net-metering rate schedule example.

Mary Cook wrote:...By the way we have a well pump--my husband tries to always run it when the sun is shining, which saves the energy lost when power goes into the batteries and then back out. Our water is stored in four 55-gallong plastic drums upstairs and runs by gravity feed down to a kitchen and bathroom sink.  pressure tanks take quite a lot of power I think...

Very interesting.  The OP should take note!

I'm going to have to come back and read through this thread -- it's late, and I'm too tired to have good comprehension on a technical topic like this right now.  But I did want to say that my approach is to go *around* my household systems, rather than trying to power everything.  I've bought rechargeable camping lights, fire starters, a small fan, and other odds and ends (I've been surprised at how much rechargeable stuff is available now) like little pumps for bucket showers; I also got a small compressor-run trucker-style frig/freezer (it's the size of an ice chest), and a solar-powered pump for the well.  I've got a solar-powered battery charger, two folding camping solar panels, and one regular solar panel to run the well pump and the small frig.  I have one battery specifically for the solar setup, a battery with my solar fence charger, and plan to get one or two more marine batteries.  I'm not making any attempt at all to power my regular household stuff with solar (and I have backups for everything, because it's cloudy a lot here).  I just figured out our essential needs -- water, light, keeping some food cool, a fan in the summer, and keeping at least a phone charged for communications/information/entertainment -- and that's all we'll power.  I have other ways to cook and heat the house.  I'm not hooking into the house wiring at all, which makes things much simpler (and a lot of this just plugs directly together, like the rechargeable stuff to the camping solar panels).  

Mike,

There's some possibility you will need to temper your expectations about solar generation.

As a background I setup and ran offgrid on my own house for 2 years, built my solar setup of 9 panels 300 watt, 3500watt outback inverter, 60 amp tristar mppt charge controller, built solar racks and my entire wiring setup.

First you need to figure out what sort of solar you're going to get, this can vary wildly, you need to figure winter vs summer coverage where you're planning your panels.  Having full sun exposure is VERY different from having partial.  A full straight on blast of sunlight on your panels can produce even higher than their rated wattage, 2800 watts for me, but even a branch of shade on a single string on that panel will drop you to 1200watts.  Only direct 90 degree sunlight will give you maximum solar, even full sunlight early in the morning at an angle may only produce 900-1000 watts on those same panels.  

While one person mentioned getting 600 watts on a rainy day, here in the pacific northwest on really dark and rainy days my array would generate about 180watt hours for the entire day.

The amount of sun and clear sun will literally change your solar production in a range of 1/20th your rated solar panels per day up to my personal maximum was 14kwh in a day (just over 5 solar hours).  To get more solar hours is tough as it requires tracking panels that will follow the sun straight on.   I live in the fraser valley of british columbia so that is about 1hr from vancouver, about 30 minutes north of the US border (in regards to latitude) and live on a south facing mountain.

My own solar was impeded by trees across the road and neighbor trees which would have made a modest difference, perhaps an additional solar hour per day maximum.  In the winter I would get shade on the panels in the middle of the day in the depths of winter.  Often a non-issue though in that the gloom and rain here would cause us to have months of an accumulated total of about 10 solar hours for the entire month (28kwh for an entire month total).

In addition to the potential issues with solars variance, which can be extreme ranging from 0.05-6x your rated panels depending on weather, you will also run into inverter drain.  Inverters, mine being an outback 3.5kw, drain around 40-45watts per hour so they accounted for about 1kwh/day drain just from the inverter being active, to have available AC power requires a standby power drain that is in addition to your actual appliance consumption.  All inverters will give you a similar issue and the drain is based on the maximum potential KW power of the inverter itself.

You will also want to really monitor your KWH consumption of power before you look to size your system.  While you've tested your deep freeze, deep freezes aren't the biggest drain in the house in our experience.  You've mentioned wanting to power a refrigerator.  That was our biggest unavoidable consumption.  We had purchased a brand new samsung 20cuft refrigerator about 6 months prior to going offgrid and the kwh consumption per day was 2.5kwh per day just for the fridge.  I would highly suggest using the kilowatt on your fridge for a few days and calculating your power consumption.

Another big issue for us was furnace consumption.  The forced air blower on our furnace (and the glowplug for startup) would consume 650 watts.  When calculating our furnace use, our furnace would run for 3.5-4hrs per day in the winter time.  That was another possible 2.8kwh per day of consumption.

I definitely advise, before spending money or sizing a system, run your kilowatt on your chosen appliances, if you can, get one on your furnace as well if thats whats intended.

What I ended up doing, once we went back onto grid power, was to get a manual multi-circuit transfer switch, this allowed me to toggle individual house circuits between grid and solar at will, each circuit is a 3 way toggle, grid-off-solar and you can actually switch them fast enough to flip electronics between power sources uninterrupted.  This allowed me to toggle the entire house over to solar (as I had that few of circuits) at will when there were power outages, and I routinely ran about 1/2 the house on the system, when there was poor solar for awhile and the batteries got lower in charge, I would just switch those circuits back to grid power and shut down the inverter until the solar charged up the batteries again.

If you do get batteries (which I did).  Lead acid are cheaper by far of various options but have a very poor charging profile that limits the potential of solar dramatically.  Lead acid batteries will have very poor life the lower you discharge the batteries, so much so that by the time you get down to 50% charge most inverters will be automatically powering off to prevent battery damage.  Lead acid take a long time to charge too as once they reach 85% charge their charge rate dramatically slows as impedance builds in the batteries.  Float charging can generally take longer than you will ever get a sustained burst of solar activity so often your batteries will not get fully charged unless you have consecutive sustained days of good sun and low consumption.  Also the round trip efficiency can end up requiring alot more power than you get back out of the batteries, ie: you use 5kwh of power, it often will require 6-6.5kwh or more to recharge that battery bank.

Lithium batteries (lifepo4 being a good choice) typically run about 4x the price per kwh but can FULLY charge in 2 hrs if you have the solar panels for it, compare that to lead acids 5-6hrs and you can start to see the benefits.  Lifepo4 batteries will often be rated for 2000-10,000 cycles too as opposed to lead acids often being 500-800.  These cycle lengths are how many times they can be reliably charged, lower discharges over their life can extend these lifetimes quite a bit as well (in both cases).  You also get a stronger charge efficiency, 5kwh of consumption may only require 5.5kwh to replace and can be replaced faster with a larger array (up to 2hr charge time from an 80% discharge).

As for wiring the panels.  Typically you will only wire 3 panels at a time in series.  The majority of charge controllers are rated for a max voltage of around 120 volts.  You CAN find higher but you need to make sure you check what you're doing.  Those are rated maximums and you are typically trying to make sure you have a safety margin underneath that, many panels are around 10 amp 30 volt but can go higher based on weather and solar (voltage wise).  Higher voltage transfers easier, as you add amps you cause impedance, getting higher voltage rather than amps lets you use lower gauge wire.

Use the wire calculators that were mentioned for figuring it out.  3 panels in series will still be 10 amps but 90 volts.  One thing I did was get 8 gauge wire and ran separate lines for each of my panel strings, around 100ft, this caused minimal loss, if I had put my strings together in parallel at the panels I would have needed at least 2 gauge or 0 gauge wire to get the same efficiency as I would have been pushing 30 amps.  The 8 gauge wire, even the 6x 100' lengths was VASTLY cheaper than 2x 100 2 gauge wire would have been.  I brought the strings into parallel where my battery bank and charge controller were.

If you go lithium batteries they have different temperature tolerances than lead acid, do your research on temperatures first based on your battery tech and it might affect where you put your batteries and charge controller.  The sooner you can get to AC power the easier it is to transfer that power with lower loss hence lower gauge wire.  For temperatures lead acid prefers around 25 celcius which is around 77 Fahrenheit for optimal use.  The lithiums are more tolerant of ranges but have serious issues below 30 Fahrenheit.

Overall solar is a good experience but you do need to temper your expectations of power.  I often said, if it only rained for 5 out of 7 days here and we only had 2 full days of sun a week that we could do solar GREAT, but the reality was that we would sometimes get WEEKS of poor weather, or entire months without seeing the sun.  Those weeks of interminable rain can result in your solar array producing much less than your inverter drains every day leading to constant net loss of power before you've even turned on a lightbulb.

The major benefit though is the availability of power, once you have that battery bank and charge controller and generator (the inverter actually is what charges the battery bank from a generator because you're bringing in AC power for that), then you have ALOT of freedom.

We managed to optimize our house to be able to sustainably survive on 2.8kwh/day with freezers, lights and refrigeration and heat.  Our 20kwh lead acid battery bank (which only lets you use about 10kwh) would last us for around 4 days without any solar.  5 Hours of charge with our propane fueled generator would bring us back up to around 95% battery and enough to last another 4 days.  If we managed this well it would run about $25-$30 in propane fuel to run entirely on generator and battery banks.  This opposed to the $30/day it would cost to run entirely on generator.  This is because your generator will consume about 40% of its maximum fuel use under load, just idling.  So if you fire up a generator, make maximum use of it otherwise you're wasting fuel.

I have plenty of other tips as well, things like sizing your system based on your batteries and necessary consumption and other tips if you're interested.

One note as well.  Go with a 48v setup.  Most charge controllers are AMP based, ie: 40 amp, 60 amp etc.  They're still an expensive piece of hardware for an mppt charge controller.  If you decide to use a 12v setup your 60 amp charge controller will only output a max of 720 watts.  It is limited to 60 amps, whether thats a 48v setup (2880 watts) or a 12v setup (720 watts).

I made that mistake early on and had to address it.  The main issue isn't the charge controller in that scenario, its the inverter.  Inverters are hard set to a specific input voltage range so a 12v inverter can't run off a 24v system or a 48v system.  A 48v inverter won't run off a 12v system.  Your good pure sine inverter is going to be the next most expensive piece of hardware after your battery bank so this is important early in the design decision.

The voltage has nothing to do with your solar array setup and just has to do with how you setup your battery array.  Stick with 48v though, can use lighter gauge wire for your battery bank and various wiring there than if you use 12v, you're also using lower amps which is generally better for heat and losses etc.

If you're sizing your array, realize that there's a maximum charging speed for any battery array.  If you are running lead acid, their maximum charge speed is around c/8 (for trojan t-105, each battery you look at should have charge specs in their spec sheets) which is 1/8th their capacity.  Lead acid slow as they get close to full but this gets you in the proper ballpark.  That means having a 5kw solar array when you have a 20kw battery bank is useless.  Your battery bank can only charge at a maximum of 2600 watts, the 5kw solar array is way under used.  (You also would need a much bigger charge controller to make use of it, but with a 20kw battery bank you couldn't charge that fast anyways).

The amount of battery storage will absolutely dictate how big a solar array you should look at.  While there is some argument for over-paneling your setup in that you produce more in offtimes, there's issues with overloading your charge controller, not being able to use that power, having to deal with dump power or other issues to get rid of excess solar electricity and possibly blowing fuses and spending alot of money in infrastructure and racking and cabling for something you fundamentally have no way to properly utilize.

If you look at a lithium battery array you can definitely look at a larger solar setup for similar capacity.  A 10kw lithium array can use between 8kw-10kw (documentation is sketchy so I'm uncertain if you can actually consume the 100% capacity they claim).  Regardless you can charge at c/2 for most lifepo4 setups which would mean 5kw per hour.  This means you would have potentially similar capacity as a 20kw lead acid but you could use a solar array 2x the size and charge the entire thing in 2hrs of sun (although you would need a bigger or multiple charge controllers).

Charge controllers can be paralleled into battery banks without issues as well.  Expanding your system is pretty easy, the thing to avoid though is mixing old and new lead acid battery banks as they have to all be ran in parallel to each other (in 48v strings) and capacity differences in different battery ages will cause total capacity disparities and can lead to the banks slowly draining or unequally charging causing discharge drains.  Adding more panels can be done in strings, you need to keep your inputting voltages similar.  I have 4x 265watt panels and 5x 300 watt panels.  Basically your voltages will align with your lowest string voltages which means if you add higher voltage panels later on, you'll end up just basically running them as though they're the lower voltage of the originals.

The inverters you go with often will allow twinning or stacking, check the literature on your brand, I know siemens inverters that someone mentioned can be paralleled or stacked into I think 3 or 4 together to increase capacity and phase your power to 240v which is good for welding and other high voltage draws.

Increasing your system size though pretty much all rotates around your batteries though, having a massive array is useless if you can't charge your batteries that fast anyways.  Its a different scenario if you're grid tied of course, you basically trade the independence and reliability of your own power and the cost of the battery system, to the utility to essentially replace a battery system with the utility (and make some possible money at the same time).  The exchange (here in BC anyways) is that our utility company has legal authority to come in and do as they please with your system and its an absolute requirement that your system go offline when the utility power is down (so you're not dangerously backfeeding into the lines when linemen may be working on the lines).  This means that your power is out along with everyone elses when it goes down.

Lots to think about I hope there's been some helpful info in here.

A helpful utility for android cell phones:

https://play.google.com/store/apps/details?id=com.andymstone.sunposition

That is a solar tracker app that lets you set a calendar day and uses your phone and location and the camera to show you the trajectory of the sun on a specific calendar day.  This can be used to figure out your solar coverage in specific areas to determine what sort of solar exposure you will get which can vary alot based on time of year.