Help sizing off grid battery bank and panels

[LRS123]

Hi all,

looking at purchasing a home that happens to be off grid with very little chance of going on grid in the future due to distance from existing utilities. Am new to solar and batteries but have been reading like crazy to try and understand what I would potentially be getting in to.

So, started out with a daily load estimate. Right now that's looking like 3500-4000 watt-hours per day including a fudge factor of 1.5. This covers lights (currently wired for D.C.), refrigerator, outlets, well pump, and washer. Dryer and hot water will be gas, but have included some watt-hours for those too.

Hand originally thought a smaller LiFePo4 battery bank was the way to go, but more I read the more I lean towards FLA. It's drawbacks are well understood and can be accounted for in design. Based on the assumptions I've come across on this site and others seem like a 48v 800-900ah bank would be appropriate. Wired in series of course.

For 800 Ah the appropriate charge rate would be ~80A, if I understand correctly. now here's where I'm a little less confident.. if I size my panel need by WH/day / hours of sun it seems like I won't be able to achieve the appropriate charge rate which sounds like it's no good. How do I size the panels correctly? My thought was charge rate x system voltage. Does that make any sense? In this case I'd end up around 4000 watts of panel. 80-90A x 48v. Do I need any additional fudge factor to account for power loss in the CC/battery/wiring?

Hopefully I'm not too far off the mark. Appreciate any any input you all have.

[SunEagle]

Hello LRS123 and welcome to Solar Panel Talk

If you need 80amps of charging then you can determine your panel wattage by looking at the max input wattage of your 80 amp CC for a 48v battery bank.

A quick calculation would be 80a x 48v = 3840 watts so that 4000 watt figure is pretty close to what it will accept.

I think going with an FLA type battery system for your first is a good choice. Since most people kill off their first bank I would not spend high dollars unless you can afford it. Just be aware that based on some input from other off grid people it will cost about $1500 for each kWh your system can generate in a day so 4000 watt hours may be around $6000 or more depending on the equipment and location of your system.

[LRS123]

Thanks for the response SunEagle. I was tossing around the idea of building a smaller/cheaper battery bank to learn on. Maybe half the capacity? Seems like I could do that pretty cheap. Would mean more generator time in the winter, but I'd rather do that than drop a ton of $ just do do it all again in a couple years. Problem is I'd like to buy all the panels now so that I can expand down the road and not worry about finding the same panel ina few years (I'm assuming mixing and matching panels is a no no). I can see us getting buy on pretty low energy use for the next couple years, but it increasing a bit beyond that as kiddos come into the picture. Any ideas as to how to build for future expansion? I guess I could always buy all the panels now and just not install them until I need to increase capacity as long as the charge controler and inverter are sized to allow for growth.

One other question: I've gathered there is some kind of ratio needed between panel voltage and battery voltage to ensure proper charging? It seems like it's necessary to use strings of 3 panels to achieve this for a 48v battery bank, but you need to make sure that the string doesn't exceed the CC voltage rating. Is that correct? I've played around with the charge controller sizing tool on midnite's site and appears panels in the low to mid 300w range work the best in strings of 3. I'm thinking either 12 or 15 panels would be necessary to achieve the total wattage. 4x3 or 5x3

[Ampster]

I think going with an FLA type battery system for your first is a good choice. Since most people kill off their first bank I would not spend high dollars unless you can afford it. Just be aware that based on some input from other off grid people it will cost about $1500 for each kWh your system can generate in a day so 4000 watt hours may be around $6000 or more depending on the equipment and location of your system.

Wow, that is a lot more that SimplPhi lithium. And they are supposed to be bullet proof. Somewhere else on this forum someone posted that he got Discovery integrated Lithium for $750 per kWhr from NAWS. I mess around with used Lithiums for $100 to $200 a kWhr but I have to add my own BMS and program all the settings to keep them healthy.

I too would recommend going with FLA for your first battery bank because you will probably kill them like I did. However if you are going to spend that kind of money per kWhr, I would recommend one of those packaged Lithium sets that are guaranteed for 10 years and are almost foolproof because you can't over charge them, they tolerate lower levels of charge without being destroyed and they are more efficient. By more efficient, I mean you don't have to oversize your panels because they don't need to be floated, they can be charged to 80% with no worries about sulfating or self discharge.
My apologies if I missed some detail about that $1,500 per kWhr metric.

[LRS123]

I think SunEagle must have been referring to the panel and associated equipment costs. I'm seeing FLA more around $150-200/KWH

[SunEagle]

I think SunEagle must have been referring to the panel and associated equipment costs. I'm seeing FLA more around $150-200/KWH

You are correct. Once you figure in the cost of the panels, racking, charge control, cabling , inverter and the batteries you can expect the amount to start to climb to that $1500 for each kWh the system can produce.

Like an idiot I purchased low wattage panels (80 & 90w) which cost about $2/watt, along with 4 AGM 12v 50ah batteries and a cheap PWM CC for my small system. Looking back I could have gotten a couple of 200w panels for $1/watt along with a couple of 6v 215ah FLA batteries and a 30amp MPPT CC. I spent about $2500 for the initial system and it can produce about 700watt hours a day but going with the second set of equipment I would have saved about $1000 and gotten a system the could produce a little more. Live and learn.

[littleharbor]

One other question: I've gathered there is some kind of ratio needed between panel voltage and battery voltage to ensure proper charging? It seems like it's necessary to use strings of 3 panels to achieve this for a 48v battery bank, but you need to make sure that the string doesn't exceed the CC voltage rating. Is that correct? I've played around with the charge controller sizing tool on midnite's site and appears panels in the low to mid 300w range work the best in strings of 3. I'm thinking either 12 or 15 panels would be necessary to achieve the total wattage. 4x3 or 5x3

One thing Midnite Solar has going for it is their Hyper VOC feature which is more forgiving than other controllers with high VOC. It is especially forgiving with 48 volt systems

If you go with 60 cell panels in series strings of three panels you should be safe with any 150 max VOC controller.

[Ampster]

I think SunEagle must have been referring to the panel and associated equipment costs. I'm seeing FLA more around $150-200/KWH

That makes more sense. Sometimes when I am traveling and confined to responses on my smartphone I am not able to see the entire context. That may be a good metric for a complete off grid system.

Despite its cost and unreliability I should be thankful for the grid and my grid tie system that cost a lot less. Based on average daily production in September (equinox) my grid tie system cost $500 per kWhr.

[LRS123]

Thanks everyone.

Ampster - we would be sticking to the grid too but it looks like we can buy a lot more house/land in this case, even with the additional costs of the system. I think given the opportunity we would do a grid tied system, but we're not the ones that are going to pay to run power 2 miles down the road to connect.

The house currently has a 20 year old, 750w array that runs the lights. It's on a 10 panel pole mount, hoping to be able to reuse the mount to save a little money.

[LRS123]

Does anyone have any thoughts on battery banks beyond the first "learner" one? It almost seems like it's cheaper to build a smaller bank with a mid quality battery and replace every 2-3 years rather than dropping the $ on a properly sized bank w/ high quality batteries? Shipping cost may add up, and I'm sure install is a bit of a pain plus added gas/maintenance for generator with the smaller bank.

I'm looking at Crown 6v 430ah vs Surrette 6v 853Ah. Assuming they last the full warranty period you'd get 5 years on the Surrette vs 6 years for 2 sets of the Crown, except the crown would only cost you 2/3 as much.

[extrafu]

You can't really compare these two batteries. The Rolls 6 CS 25P (853Ah) is a top of the line battery - the Crown is average. I do own 8 x 6 CS 25P batteries - just replaced my Rolls S550 (similar to the Crown you're looking for). The 6 CS 25P will not suffer from the same voltage sag the Crown will be. If you discharge them slowly, they (C/50 - C/100 - which is still *a lot of power* on a 48V system), they will last forever. You also get 4250 cycles at 50% DoD.

[LRS123]

Thanks extrafu. I've been tracking our current power use and it does seem like we often are drawing in that 50-75ah range.

Can you explain the the effects of voltage sag? I'm assuming this just refers to the voltage drop as DoD increases?

So in your mind the cost is justified? Once you are comfortable with the technology at least.

[Mike90250]

I'd say use a set of beater batteries, and if you get the length of the warranty out of them, then go for the expensive ones.

I might be of the mind of getting the cheap batteries and not worrying so much about getting 6+ years out of expensive ones. every 2 or 3 years, fresh batteries !

[extrafu]

@LRS123 You have to finalize your estimates about power usage. Also, note that 75A @ 48V is a lot of power. The Crown batteries you're looking at will likely suffer a 2V voltage drop when you suck as much as power. The Rolls 6 CS 25P - probably 0.5-0.8V from my own observations. That has impacts on a lot of stuff, for example - how your auto-generator start is configured when voltage is too low, and for how long. How your inverter is configured to shut down when voltage is too low, and for how long.

If you plan to suck 50-75A once or twice per day for a minute, I wouldn't worry too much. If it happens 20 times per day, that's an other story.

[LRS123]

extrafu - I have a kill-a-watt meter that should show up today so I can fine tune things. It will still be just an estimate as not all the appliances in our current home will carry over to this new one, but its the best I can do.

The big current demands I see are:
the well pump, likely a 120v Grundfos 10gpm or similar - ~30 amps on a 48v battery
fridge/freezer - likely a chest style for both - assume starting amps are pretty high for each, maybe 30A range?
washing machine - again starting amps are probably high - 20A range
Microwave/vacuum - 25 A range

Worst case scenario is well pump, fridge/freezer all come on while using one of the other high load devices. That's a lot, but hopefully it would be several min max. Thanks for pointing this out. I'd only been thinking of it from the perspective of it the inverter could handle it, not necessarily the demand on the battery bank.