Starting from nothing, but going big

TRW said:

I think you have a great point. I wonder if many just are not used to the concept of full independent power, without the grid. The cost of a storage system continues to fall, and my only reservation is that the life of the batteries is a bit limited, so the cost is still somewhat high.

Click to expand...

Q-Dog said:

If I were starting from scratch I would build a new, super-efficient house with good southern roof exposure, then size a system for the new house. Nothing about my house is conducive to being powered by solar ... too little insulation, wrong roof angles, too many trees for ground mounting. It would be cheaper to start from a clean slate.

Click to expand...

GXMnow said:

Here is my $0.02

If the grid is there, have it available. A minimal connect fee and setup for zero export can be pretty cheap in most areas. And if you do fall short from solar, you don't have to fire up a generator.

I totally agree you need to do a basic energy audit to get a good idea how much peak power you need, and how many watt hours you need each day. If the budget is there, I would have 2 days of battery storage for the basic needs. I just had 2 days in a row where my solar produced only 1/5 of a normal day output. Obviously I was having to use grid power. Batteries are still expensive enough, that cutting it close and having the grid available is going to be a lot cheaper. Depending on where you are located could also make a big difference. How many miles a day do you drive your electric car? Will it be able to charge near solar noon, or will it charge off of your battery bank at night? Do you have city water, or do you need to run a deep well pump? All of my heating is natural gas. I can't run my central A/C on my solar/backup system. For an off grid home, I would use a mini split in each main living area. Many of them can do heat pump heating as well. That is far more efficient than a resistance heater.

The battery bank is the big one. I currently have just 18 KWH of Li NMC battery which is about 30% of a Chevy Bolt EV pack. I am working on a deal to double that up. When these need to be replaced, I will see where technology has gone. Are you comfortable with a DIY battery bank or do you want factory built with a UL logo? "Vehicle to Home" bidirectional charging is coming, and is actually here, if your car is a Nissan Leaf. The charger can take your home AC power and charge the car, obviously, but it can also reverse and take power from the car's battery and push it into the home AC system and give you backup power. The one system on the market now is pretty limited and expensive, but I see this getting big soon. If you have a car with an 80 KWH battery pack, you have more storage than most of us on here. Hopefully they will get this for CSS DC fast charge ports soon. And Tesla is off on their own.

As for the equipment, there are a bunch of choices based on your power needs. If I was building from scratch, I would still go with enough Enphase microinverters to supply all my normal AC power loads that are in use while the sun is up. I have 16 x 300 watt panels on 240 watt iQ7 inverters. At full sun, they top out at making 3,900 watts into my home. That is more than my normal peek loads, except for the A/C in the summer. I get up to 7 Sun Hours and top 30 KWHs on good days. I do have the rare cloudy 6 KWH days though. On a new system, I might trim that down by 2 to 4 panels. Then have another 12 panels DC coupled through charge controllers to charge the battery bank. The battery inverter choice is a harder decision. I have a Schneider XW-Pro and I have to admit, the hardware is excellent and it is a powerful rock solid beast. BUT... Their software has a few holes in it and their tech support is a little weak as well. But my biggest complaint is mostly fixed with the addition of the DC coupled solar panels. With just AC coupled like I have, the system does not time shift power on it's own. But off grid it seems to work much better. Running in a zero grid export mode, and having DC coupled solar it is hard to beat. The Outback Skybox has some very good features and combines several functions into a single box solution. But the battery inverter is only rated at 5,000 watts, and can have issues on unbalanced loads and starting surges. Many people are raving about the Sol-Ark, and it does look pretty good, but again, it works it's best with DC coupled solar. As much as I have complained and moaned about the issues with the Schneider software, I still think it is a great product for a battery inverter. My gripe is all about being grid tied and only having AC coupled solar. Schneider also does the nickel and dime thing. Each major feature is another box to buy. When I add the DC solar, it is a $500 charge controller to add just 2,400 watts of panels. If I want true zero export, that is over $300 to add a Watt Node box. Those are both built in the Skybox.

Work up an estimate of your true power needs and let's figure out what you need.

Click to expand...

Hedges said:

I can buy PV and GT inverters for about $0.50/watt. With rack mounts and electrical stuff, easily under $1.00/watt. (my labor is free)
Amortized over 10 years I put that at $0.05/kWh. Replace an inverter at that point, and over 20 years it amortizes to $0.03/kWh.

Batteries are more expensive than that, but getting cheaper. If I wanted 3650 deep cycles over a decade, I would do DIY LiFePO4.
15 months ago I went with AGM because I don't anticipate even 365 power failures and deep cycles (grid-backup system.) The only alternative I knew of was commercial lithium, longer life an proportionately higher price. But DIY LiFePO4 is lower cost than AGM. Biggest issue is it needs BMS able to supply the highest surge current you need.

I use Sunny Boy GT inverters and Sunny Island battery inverters, and would do the same today (if Sunny Island still available for 25% to 50% of msrp.)
But I don't think this system can do peak shaving, store PV generated power during hours of low rate and then backfeed the grid from battery during high rates. Other products have the firmware to do that.

I also like to over-panel PV to about 150% of GT inverter (or SCC) capacity, using multiple orientations. This gets more Wh/day out of the same electronics.

I have a small battery, just enough to make it through one night. During the day, if off-grid, PV power is use-it-or-lose it. This is a good way to avoid putting money into batteries. If you need several days of autonomy, a larger battery plus an auto-start generator might be the way to go. People here have put diesel generator power at $0.50/kWh. You have to compare that to cost of battery, considering how often you need more than 12 or 24 hours worth of power to come from the battery.

Click to expand...

TRW said:

I enjoy your input and have read much of what you have written over the last year. You have lots of good info here, so lets pretend you won the lottery and are building a new house, and want a new solar system complete, one that you can not use anything you have, can you list what that would be? If you are willing to offer that, then I am very curious where you would buy it. I will ask any of you the same question?

Click to expand...

Hedges said:

Have to consider whether it is on-grid or off-grid, and if on-grid are utility rates time-of-use?

Good net metering is better and cheaper than the best and cheapest battery. Use that if you can.
I wouldn't use lithium polymer, because even UL listed units have caught fire. Or if I did, located in a separate bunkhouse.
DIY LiFePO4 looks attractive (low price and high cycle life), with cells having welded on studs, and a BMS capable of delivering surge current. Or bypassed with a relay.
The AGM I presently use works fine for high surge current. It wouldn't be good for frequent deep cycling.

If time-of-use rates, it would be good to store PV generated power prior to 4:00 PM and spin the meter backwards 4:00 to 9:00 PM when rates are 3x as high.
I don't have experience with any inverters that do peak shifting. My Sunny Island (23kW continuous 44kW surge from 4 inverters) don't support that. Sunny Boy Storage does, but as a back-up inverter (with separate autotransformer and transfer switch) it is only 6kW continuous, 7.7kW maybe 9kW surge.
For peak shifting, I'd have to look over various companies and products, find something which also supports surge loads to start motors.
But when I looked at my cost for GT PV vs my cost for batteries, I think it is better to backfeed the grid with 3x the kW in the morning rather than buying a battery to store 1x the kW to be delivered during peak rates.

Rather than just an array facing South, I would have arrays facing SE and SW, so power production is spread over the day. This reduces peak battery charge current or grid export and gives more hours of production, for about 40% more Wh/day.

I would consider 3-phase. With my present split-phase utility connection, I could have only 6.7kW come from grid or backfeed grid through the 56A relay of one 120V Sunny Island. The other two would create the missing phases. With 3-phase connection 3x as much power could pass through, or off-grid 6x as much PV power could be supported. What I have now is split-phase with four Sunny Island, so 26 kW can pass through.

PV panels - find one known to be reliable. Accelerated stress testing shows high failure rates for some brands, as has field experience. When I added on last year one I considered was SunPower. Nothing I read questioned reliability, only whether it was worth the higher price. There was also that claim that of N-type silicon doped to make a cell vs. P-type, one has lower degradation rate, and that's what SunPower used. I'm not certain about the claims, but went with that brand. There was one model "P17" series I considered but it had some new and unusual aspects to how it was built so I bought more conventional "E20" series instead.

Software/firmware - I don't want anything updated. The boxes should work together as delivered and perform the functions specified in the manual. Shouldn't have to push updates to get the to do what was originally promised. Maybe a manual update to change from UL-1741 to UL-1741SA if desired.

Internet connectivity - Don't want to be a victim of Stuxnet, Sandworm, or the like. Nothing from the outside should be able to change the system unless I authorize it. A method to push data would be OK if there are no vulnerabilities, but I don't trust software, especially networking.

For the functions it does implement, my Sunny Island/Sunny Boy system works fine. I would use older transformer type inverters not the newer transformerless, because I think they're better at filtering out poor power-factor from switching power supplies like a VFD I have.

I have wondered if Sunny Boy Storage would work on a Sunny Island system, responding to frequency-watts and storing/providing power when needed. That would keep AGM batteries on Sunny Island always floating, except when they supply surges. Sunny Boy Storage is transformerless, and at least some of the 400V batteries compatible with it are lithium polymer. REC offers BMS for high cell count batteries; that might work to make a 128s LiFePO4 battery (100 kWh) for Sunny Boy Storage.

I know there are other good brands such as Schneider, Outback, Victron, but I don't have experience with them. I might consider if they had features I needed. I would look for units that could be stacked for higher wattage.

Where to buy - used/overstock vendors like SanTan have good deals on PV panels compared to retail outlets. There are probably other vendors who are good but I have bought from him.
For inverters and batteries, I've used Google to find listings, also eBay. Many retailers use eBay as a platform to advertise and it is the first place I look for everything. (But for LiFePO4 cells, forum members have identified good sources.)
I'm obviously not looking for a lot of support - I'm an EE and a contractor. For the inverters I've used SMA's help system.
Some of the retailers can put together a package and design the series/parallel configuration. AltE store for instance, who is represented on this forum.

Click to expand...

Hedges said:

Have to consider whether it is on-grid or off-grid, and if on-grid are utility rates time-of-use?

Good net metering is better and cheaper than the best and cheapest battery. Use that if you can.
I wouldn't use lithium polymer, because even UL listed units have caught fire. Or if I did, located in a separate bunkhouse.
DIY LiFePO4 looks attractive (low price and high cycle life), with cells having welded on studs, and a BMS capable of delivering surge current. Or bypassed with a relay.
The AGM I presently use works fine for high surge current. It wouldn't be good for frequent deep cycling.

If time-of-use rates, it would be good to store PV generated power prior to 4:00 PM and spin the meter backwards 4:00 to 9:00 PM when rates are 3x as high.
I don't have experience with any inverters that do peak shifting. My Sunny Island (23kW continuous 44kW surge from 4 inverters) don't support that. Sunny Boy Storage does, but as a back-up inverter (with separate autotransformer and transfer switch) it is only 6kW continuous, 7.7kW maybe 9kW surge.
For peak shifting, I'd have to look over various companies and products, find something which also supports surge loads to start motors.
But when I looked at my cost for GT PV vs my cost for batteries, I think it is better to backfeed the grid with 3x the kW in the morning rather than buying a battery to store 1x the kW to be delivered during peak rates.

Rather than just an array facing South, I would have arrays facing SE and SW, so power production is spread over the day. This reduces peak battery charge current or grid export and gives more hours of production, for about 40% more Wh/day.

I would consider 3-phase. With my present split-phase utility connection, I could have only 6.7kW come from grid or backfeed grid through the 56A relay of one 120V Sunny Island. The other two would create the missing phases. With 3-phase connection 3x as much power could pass through, or off-grid 6x as much PV power could be supported. What I have now is split-phase with four Sunny Island, so 26 kW can pass through.

PV panels - find one known to be reliable. Accelerated stress testing shows high failure rates for some brands, as has field experience. When I added on last year one I considered was SunPower. Nothing I read questioned reliability, only whether it was worth the higher price. There was also that claim that of N-type silicon doped to make a cell vs. P-type, one has lower degradation rate, and that's what SunPower used. I'm not certain about the claims, but went with that brand. There was one model "P17" series I considered but it had some new and unusual aspects to how it was built so I bought more conventional "E20" series instead.

Software/firmware - I don't want anything updated. The boxes should work together as delivered and perform the functions specified in the manual. Shouldn't have to push updates to get the to do what was originally promised. Maybe a manual update to change from UL-1741 to UL-1741SA if desired.

Internet connectivity - Don't want to be a victim of Stuxnet, Sandworm, or the like. Nothing from the outside should be able to change the system unless I authorize it. A method to push data would be OK if there are no vulnerabilities, but I don't trust software, especially networking.

For the functions it does implement, my Sunny Island/Sunny Boy system works fine. I would use older transformer type inverters not the newer transformerless, because I think they're better at filtering out poor power-factor from switching power supplies like a VFD I have.

I have wondered if Sunny Boy Storage would work on a Sunny Island system, responding to frequency-watts and storing/providing power when needed. That would keep AGM batteries on Sunny Island always floating, except when they supply surges. Sunny Boy Storage is transformerless, and at least some of the 400V batteries compatible with it are lithium polymer. REC offers BMS for high cell count batteries; that might work to make a 128s LiFePO4 battery (100 kWh) for Sunny Boy Storage.

I know there are other good brands such as Schneider, Outback, Victron, but I don't have experience with them. I might consider if they had features I needed. I would look for units that could be stacked for higher wattage.

Where to buy - used/overstock vendors like SanTan have good deals on PV panels compared to retail outlets. There are probably other vendors who are good but I have bought from him.
For inverters and batteries, I've used Google to find listings, also eBay. Many retailers use eBay as a platform to advertise and it is the first place I look for everything. (But for LiFePO4 cells, forum members have identified good sources.)
I'm obviously not looking for a lot of support - I'm an EE and a contractor. For the inverters I've used SMA's help system.
Some of the retailers can put together a package and design the series/parallel configuration. AltE store for instance, who is represented on this forum.

Click to expand...

Hedges said:

If you had 10kW total panels, 5kW each direction with a 90 degree angle between them, trigonometry or A^2 + B^2 = C^2 says the diagonal is 0.7 as long as sum of two sides, the two arrays would present 7kW worth of area toward the sun at Noon.

Click to expand...

Hedges said:

Typically arrays have been oriented one direction, South with optimal tilt if possible, to produce the most power throughout the day.
Some homes have multiple orientations because one roof area wasn't large enough.

There have been mechanical tracking mounts, but they are expensive and can break down. PV panels have become so cheap that if you have the area you can just put up more panels facing South to get the extra watt hours a tracker would have produced (but power will peak middle of the day, taper off in the evening when you might still want to run A/C, so it would start to draw from batteries or grid.)

My idea is, instead of having an array that swings from SE to S to SW during the day, just make an array aimed SE and another aimed SW. If you had 10kW total panels, 5kW each direction with a 90 degree angle between them, trigonometry or A^2 + B^2 = C^2 says the diagonal is 0.7 as long as sum of two sides, the two arrays would present 7kW worth of area toward the sun at Noon. But more area morning and evening than a flat array. That is, you can put 40% more panels on the inverter, and their production will be spread longer over the day.

There is some penalty, less watt-hours from a panel aimed SW than due South. But fitting more panels on the existing inverter/charge controller and have more uniform production during the day seems worthwhile.

If your only roof slopes are facing North and South, then probably just cover the South face and don't worry about SE and SW.

Panel wear-out could be from visible light, from UV (can darken epoxy), maybe heat affecting the doped silicon, heat damaging backsheet, water intrusion. Quality panels last for decades, but some became defective within a few years.

Click to expand...

​

Hedges said:

Have to consider whether it is on-grid or off-grid, and if on-grid are utility rates time-of-use?

Good net metering is better and cheaper than the best and cheapest battery. Use that if you can.
I wouldn't use lithium polymer, because even UL listed units have caught fire. Or if I did, located in a separate bunkhouse.
DIY LiFePO4 looks attractive (low price and high cycle life), with cells having welded on studs, and a BMS capable of delivering surge current. Or bypassed with a relay.
The AGM I presently use works fine for high surge current. It wouldn't be good for frequent deep cycling.

If time-of-use rates, it would be good to store PV generated power prior to 4:00 PM and spin the meter backwards 4:00 to 9:00 PM when rates are 3x as high.
I don't have experience with any inverters that do peak shifting. My Sunny Island (23kW continuous 44kW surge from 4 inverters) don't support that. Sunny Boy Storage does, but as a back-up inverter (with separate autotransformer and transfer switch) it is only 6kW continuous, 7.7kW maybe 9kW surge.
For peak shifting, I'd have to look over various companies and products, find something which also supports surge loads to start motors.
But when I looked at my cost for GT PV vs my cost for batteries, I think it is better to backfeed the grid with 3x the kW in the morning rather than buying a battery to store 1x the kW to be delivered during peak rates.

Rather than just an array facing South, I would have arrays facing SE and SW, so power production is spread over the day. This reduces peak battery charge current or grid export and gives more hours of production, for about 40% more Wh/day.

I would consider 3-phase. With my present split-phase utility connection, I could have only 6.7kW come from grid or backfeed grid through the 56A relay of one 120V Sunny Island. The other two would create the missing phases. With 3-phase connection 3x as much power could pass through, or off-grid 6x as much PV power could be supported. What I have now is split-phase with four Sunny Island, so 26 kW can pass through.

PV panels - find one known to be reliable. Accelerated stress testing shows high failure rates for some brands, as has field experience. When I added on last year one I considered was SunPower. Nothing I read questioned reliability, only whether it was worth the higher price. There was also that claim that of N-type silicon doped to make a cell vs. P-type, one has lower degradation rate, and that's what SunPower used. I'm not certain about the claims, but went with that brand. There was one model "P17" series I considered but it had some new and unusual aspects to how it was built so I bought more conventional "E20" series instead.

Software/firmware - I don't want anything updated. The boxes should work together as delivered and perform the functions specified in the manual. Shouldn't have to push updates to get the to do what was originally promised. Maybe a manual update to change from UL-1741 to UL-1741SA if desired.

Internet connectivity - Don't want to be a victim of Stuxnet, Sandworm, or the like. Nothing from the outside should be able to change the system unless I authorize it. A method to push data would be OK if there are no vulnerabilities, but I don't trust software, especially networking.

For the functions it does implement, my Sunny Island/Sunny Boy system works fine. I would use older transformer type inverters not the newer transformerless, because I think they're better at filtering out poor power-factor from switching power supplies like a VFD I have.

I have wondered if Sunny Boy Storage would work on a Sunny Island system, responding to frequency-watts and storing/providing power when needed. That would keep AGM batteries on Sunny Island always floating, except when they supply surges. Sunny Boy Storage is transformerless, and at least some of the 400V batteries compatible with it are lithium polymer. REC offers BMS for high cell count batteries; that might work to make a 128s LiFePO4 battery (100 kWh) for Sunny Boy Storage.

I know there are other good brands such as Schneider, Outback, Victron, but I don't have experience with them. I might consider if they had features I needed. I would look for units that could be stacked for higher wattage.

Where to buy - used/overstock vendors like SanTan have good deals on PV panels compared to retail outlets. There are probably other vendors who are good but I have bought from him.
For inverters and batteries, I've used Google to find listings, also eBay. Many retailers use eBay as a platform to advertise and it is the first place I look for everything. (But for LiFePO4 cells, forum members have identified good sources.)
I'm obviously not looking for a lot of support - I'm an EE and a contractor. For the inverters I've used SMA's help system.
Some of the retailers can put together a package and design the series/parallel configuration. AltE store for instance, who is represented on this forum.

Click to expand...

chrisski said:

I like that math. Seems perfect for when the sun rises at 090 true and sets at 270 true, but how would you adjust this throughout the year as it goes more to the North or South?

Click to expand...

Hedges said:

I don't want to adjust anything. Motorized trackers didn't look like a great deal when I paid $4.00/watt for PV, certainly aren't worth buying now that it is about $0.33/watt.

For a fixed tilt, you might maximize year-round Wh for a net-metered system. You might maximize winter for off grid (or maybe summer, if your primary load is A/C.)

To optimize a two orientation setup like I propose, you could enter each into an insolation calculator and add them together.
I've only taken something of a SWAG at it. My original setup I made orientation and tilt about 2:00 PM to maximize peak rate Noon to 6:00 PM Summer production. I now have so much PV it doesn't need to be optimum.
I wasn't thinking due East and West (90 and 270) but SE and SW.
I'll leave my 2:00 PM orientation for a couple of my arrays, but two that have become shaded I might reorient toward 10:00 AM. The goal is to max out the inverters for more hours.

Of course, if you have roof angles to work with, that's probably what you use rather than tilting the panels differently.
The trick to all this is that parallel strings on one MPPT do not have to be same orientation. They can be quite different, and the penalty was found to be only 2%. So you get more use out of your electronics by overpaneling this way. (partial shading of one string on the other hand can cause a lot of power loss.)

Click to expand...

TRW said:

You offered this comment, and I like it.

Some of the retailers can put together a package and design the series/parallel configuration. AltE store for instance, who is represented on this forum.

At one time I thought that by now I would understand this all well enough to do this on my own, but I think that I may need to go with someone like the AltE store, I don't think I want to go with used panels, I may as well do it the best I can, I have talked to several people who have changed panels after 10 years, so I am hoping to skip that step. I have more to digest in your post, but I really appreciate your input and time!

Click to expand...

Horsefly said:

I'm guessing that there are thousands of different approaches to this, so I'm not surprised we differ here. What I'll say is this: I iterated several times in designing my system, and finally I realized that the battery really was the heart of the system. I really think you need to define the battery size first, and size the array based on that. If you have no cloudy days, then yeah you can size the solar array based on your energy consumption.

There is more than one way to do it, but I feel pretty strongly that the battery is the core, and you go from it to the solar, not the other way around. I've run that by quite a few experts, and a majority - maybe slim - agree with that (but maybe they just don't want to argue with me ). I think if you are building a system to provide power - not just convert solar to power - you have to do it that way.

We all have our opinions, and we know what opinions are worth!

Click to expand...

Sanwizard said:

Alt-e has very reasonable prices. I bought my panels from them. Watts247.com is also very reasonable. I bought two LV6548 inverters from them. For batteries, if you want inexpensive and powerful, buy the cells and BMS, and put it together yourself. My 26KW Lifepo4 was built for less than 4K. Off the shelf you will pay 4x as much or more, but you will not be able to call someone if something breaks. Always tradeoffs in life.
I also like the flexibility of the latest hybrid inverters. Makes things simple. Take a peek at what DanF and DMI inc are doing.

Click to expand...

mzoid said:

Thanks for this back and forth on design philosophy.
I'm at the stage where I've got a design in mind after power budgeting and considering my site, but have not yet pulled the trigger on the main items for the system yet. I've had covid, and then long covid, and the mental fog that comes with that means I really couldn't tell you which came first for me: battery size or array size. Too many late nights researching, learning organically, crunching numbers backwards and forwards... it is all a blur.

But I THINK array size came first for me, and I think this is because I'm designing for a remote location where the grid is less reliable, and where complete energy independence from fossil fuels becomes a necessary consideration, should disruption to global supply chains occur.

I think that's the difference between the two primary approaches you have both been discussing. If being attached to the grid is a prominent part of the plan and there is no concern that fuel for a generator or propane for heating/cooking might cease to be sourceable, then the battery takes precedence. You can squarely worry first and foremost about your needs and size the battery accordingly. You'll always be able to look to the grid, or to a generator to feed the batteries if the sun doesn't deliver everything you need on occasion.

When the sun really is the only energy input into the system in a worst case scenario, and you start thinking about strategies to function that way, then the array takes precedence in my opinion. Alot of night time power usage is luxury. And as you transfer as much power consumption from night into the sunlight hours so it can be provided for by the array, the size of the array becomes more important. The more you are willing to give primacy to considering complete energy independence, the more you start thinking the way humans have thought about survival since time immemorial: use what is provided when it is provided. Live with the seasons. Work while the sun shines. Sleep at night.

When the day time activities are covered by the array, then based on this lean mindset you have now enforced, you can start thinking about how many consecutive days of low solar production would start to become a serious concern. Realistically, with a big enough array, you're still going to be able to run your refrigerator during the day and charge the battery enough for it to run during the night, unless there's a volcanic eruption that obscures the skies for weeks on end. So the battery size becomes more about what your wants are, as opposed to your needs.

Most people don't think about things this way, and I suspect that would lead to the battery being given primacy most of the time. When wants are thrown into the same pile as needs, and the assumption is there will always be an energy source to feed the batteries to cater to those wants and needs, then naturally, the battery is the first consideration, and whether the panels supply all of that or just some of it, can be worried about in step 2.

Fascinating to unravel some of the assumptions behind the philosophy of these designs. Thanks everyone for sharing your journeys here.

Click to expand...

mzoid said:

But I THINK array size came first for me, and I think this is because I'm designing for a remote location where the grid is less reliable, and where complete energy independence from fossil fuels becomes a necessary consideration, should disruption to global supply chains occur.

Click to expand...

Horsefly said:

Not to beat a dead horse here, but this is THE major reason to size the battery first. Unlike some people , I'd say it isn't wise to design based on the sunniest day. I know that isn't what you are saying, but you need to realize you will have days (probably multiple) with not much sun. If you are putting the emphasis on the array, it isn't doing much for you.

Turning to some of your other points: If the grid is unavailable and you don't have fuel for the generator, what happens on those days that are not real sunny?

Also, saying you will be able to run the fridge on cloudy days with the limited sunshine might be a bit optimistic. A fridge is a "medium" load. Not as big as a well pump or microwave, but bigger than most of the other "necessary" loads.

I think that especially for a truly off grid situation (like mine), you should put the emphasis on the battery, knowing that oversizing or adding to the PV array is easy and inexpensive. That array does no good for you if you don't store it up for the nights and cloudy days.

In the end, it's up to you. Know that for suitcases, paychecks, and batteries you eventually use all you have and wish you had more.

Click to expand...