Coming to LiFe

[Coydon]

I'm looking to build my second solar station iteration. I currently have a 12V FLA 225Ah system connected to a single 350W panel by way of a 30A MPPT charger controller. It has been plenty for my basic startup needs, although I am looking to add some refrigeration to my campstead this year. I'd like to go larger and give LFP a try, keeping the older system as a backup/second location option. As I build a more permanent structure, I will not be moving toward a conventional "grid-style" power system, but will continue to use various 12V appliances with a chest freezer on an inverter, with other 120 devices only occasionally.

I found some used panels at a good price. I picked up 6 Sunpower "327W" panels, the seller said they were 3-5 years old and should be working above 95% at least, I believe. It wouldn't be too difficult to pick up more if needed, but the shed and the spot I have planned for it would not support more than 6, I'd have to relocate and run much longer wires.

Looking at likely appliances, I'm guessing my daily power demand would be around 1-1.5kWh. I have no set habits and will adjust as needed, so the primary concern is to find a 'sweet spot' in available materials to move forward with. Refrigeration is the primary anticipated need, but given the fact that days with greater need for power will coincide with higher solar input, the usual warnings I see about 'buffer' calculations seem overly alarmist for my instance. I'm picking 1.5kWh/day demand and 3 days reserve along with 2 hours of daily solar insolation in my northern clime as arbitrary points to move forward with. This will be much greater most of the year but anticipating our short, cloudy fall and winter days (including a 60* mounting angle for the panels). Keeping a healthy battery bank is more important than having a suplus of energy stored at any given time.

Seems to me a DIY bank with prismatic cells would be worth it to save a bit of money over prefabricated rack systems. I would like to have a reliable system and just focus on other things, but this is essentially paying myself to learn more about solar charging and know how to handle problems/changes in the future.

I currently have an EPEver Tracer 3215BN 30A charger, but it only does 12/24V systems. I was ready to move to a 48V system. However, I'm also calculating that the battery bank should be about 5kWh total. Pricing out prismatic cells, 16 100A cells would cost around $1000 and require a new charger purchase to use with the 48V system. OTOH, 8 300A cells would cost about the same, store 50% more power, and allow the use of my current charge controller as a 24V system. I think. If I run the panels in 2 strings of 3, it would be 18A @ 130V (Sunpower mono panels w/Max Voc of 64.9, Max A 5.98). Is that a good idea? Will the EPEver Tracer (max W 780) manage over-paneling?

What other calculations/considerations am I missing? I'm not clear on the whole C calculation thing yet, trying to digest everything in small bites. It seems to me the 1.5kWh input per day would replace my fallback draw amount, but is that enough to keep the LFP bank happy?

 

[zcskywire2]

I can really only speak on the battery side of things. The first is that with a self assembled lithium iron phosphate battery pack is that you will need a BMS. This is to protect the cells from over charging, over discharging, over temperature operation, under temperature discharging and over current conditions. You will need a properly sized bms for both for the number of cells and the power you are expecting to draw out of it in amps. Based on the worse case power draw prediction of 1500w, thats around 30 amps of draw with a 51.2v 16 cell pack, 60 for a 25.6v 8cell pack. However no inverter is 100% efficient, you should expect a higher draw than the simple math would suggest. Furthermore most BMS's with internal switching tend to over rate their actual current capabilities. It would be good to overspec the BMS by a good margin for your application as a safety measure. Furthermore this gives you more flexibility if you have to add more loads in the future.

The second thing is fuses for any serious load. Things can go very wrong very fast with the amount of power instantaneously available from a lithium battery. A fuse is a safety for a gross overload situation to prevent even more dangerous events from occuring in a fault condition. You should have at least one properly sized fuse on the pack output and possibly some on your individual loads.

On keeping the batteries happy. If you are getting net power or neutral loading over the 24 hour day than the batteries will be fine. Lithium iron phosphate batteries do not need a constant charge like lead acid chemistries to stay healthy. As long as you do not over or under charge them you will be fine.

On C rate, the C rate is a ratio between the input/output current of a battery or pack and the rated amp hours of the pack. So on your planned 51.2v 100Ah pack 1C is equivalent to 100A of discharge or charge. For the This is irregardless of the pack's or cells voltage. A higher C rate puts more wear on the battery and will reduce it's lifespan. Most cells used for home solar have a continual charge or discharge rate at 0.5C to 1C. Either one of the setups you mentioned would be well below 1C in max power consumption.

The last thing I can think of is watch out for high inrush currents, both on the devices you are running in your system and when you are setting it up. Old fridges and freezers with compressors can have a huge spike in power draw when turning on and this can overload some devices. Likewise inverters can have a big inrush currents when wired leading to large sparks. These sparks will increase in magitude as the system size and voltage goes up. It may be good to have a proper resistor to use to "pre charge" an inverter when wiring then up.

I hope this long ass mess helps with some stuff. If you have any battery questions please ask. Sadly really can't speak on solar stuff.

 

[Coydon]

You will need a properly sized bms for both for the number of cells and the power you are expecting to draw out of it in amps. Based on the worse case power draw prediction of 1500w, thats around 30 amps of draw with a 51.2v 16 cell pack, 60 for a 25.6v 8cell pack. However no inverter is 100% efficient, you should expect a higher draw than the simple math would suggest.

The '1500w' I named thinking of a string of cloudy fall/winter days where I would avoid all voluntary appliance usage and rely on the batteries only to power the refrigerator/freezer as the cumulative draw for the day. I think you are describing the peak draw at any single moment that the BMS might have to handle?

 

[zcskywire2]

I think you are describing the peak draw at any single moment that the BMS might have to handle?

I was peak continual draw the BMS would see, though peak pulsed power is as important. You should watch for both when properly sizing a BMS setup

 

[DIYrich]

You could configure the 16 cells as 8serial, 2 parallel, 24v system. That will give you 200 amps at 24v. You could use one BMS, but each "cell" the BMS sees is two cells connected in parallel. If you get 2 BMS, you could see each individual cell.

4s4p would be for a 400 amp 12v system. similar comment about BMS, but now you need 4 to see each individual cell.

Be careful that the batteries stay above freezing. Use a heater if you have to.

 

[yabert]

Seems to me a DIY bank with prismatic cells would be worth it to save a bit of money over prefabricated rack systems. I would like to have a reliable system and just focus on other things.

If you want to build a battery to learn, it's the best way and LFP cells are quite safe and excuse most errors.
If it's to save cash, think twice because of the shelf 24v or 48v battery are relativly cheap for a ''small'' battery format. Look at LiTime as example.
When you count times, bms, fuse, wiring, cables, tools and casing you could end with an higher price than of the shelf battery.

 

[Coydon]

You could configure the 16 cells as 8serial, 2 parallel, 24v system. That will give you 200 amps at 24v.

This would allow me to use my current charger, but I would end up with 2/3 the power as if I bought 8 300ah cells, and they would take up twice the space, require extra hardware, all for roughly the same battery price. Are there other benefits to having more low powered cells?

When I was first researching things last year I read various comments that you should never mix cells of different ages/usage levels in a configuration because the oldest/weakest would quickly draw down the rest to it's level. Does that not apply in DIY setups where you can monitor and attend to the status/performance of each cell carefully?

 

[Coydon]

Be careful that the batteries stay above freezing. Use a heater if you have to.

I would only be using a charger with temperature controls to prevent charging below freezing (though I'm confused after watching Will's Battleborn interview if the dangerous temps start at 40, 32, 25, a continuum...?). Is there damage that can occur from freezing besides the plates coating up during charging?

I hope to have a root cellar to place the batteries in before winter, otherwise they would be in an 'unthermostatted' shed. I will be sure to look into insulated containers when building the box. But this brings up another consideration in 24V vs 48V systems- that choice doesn't affect the distance between the panels and the charge controller/battery bank does it? My understanding is that how the panels are wired for voltage would determine demands on that wiring, and that the battery system voltage would only affect the size of wires needed to cover various distance from the bank to consumption points...?

 

[yabert]

you should never mix cells of different ages/usage levels in a configuration because...

There is no real problem to mix cells of same chemistry at the moment they are mixed in parallel.
Example:
-Mix 8 pairs of old 100Ah (so 200Ah in parallel) with 8 200Ah cells to build a 16 cells battery (51.2V 200Ah) = WRONG!
-Mix 8 old 100Ah cells with 8 200Ah cells in a configuration where each 100Ah is in parallel with each 200Ah to create a 8 cells battery (25.6V 300Ah) = Good!

 

[DIYrich]

This would allow me to use my current charger, but I would end up with 2/3 the power as if I bought 8 300ah cells, and they would take up twice the space, require extra hardware, all for roughly the same battery price. Are there other benefits to having more low powered cells?

300 ah cells are bigger and heavier. A 100ah cell is 130.3mm x 36.7mm x 200.5mm. A 280ah cell is 173.7mm x 72.0 mm x 207.5mm. Don't know the dimensions of a 300 ah cell. 1980g vs 5420g.

I'm all for cheapest cost per watt. You asked about 16 requiring a 48v system. I merely showed you that 16 cells can be configured for 12v or 24v too.

 

[zcskywire2]

The 302-315 ah cells tend to match the 280ah in size. Only outlier is the CATL 310 which is slightly taller. Weight wise all the larger capacity cells are heavier

 

[Coydon]

Are there known differences of relvance between CATL, CALB, and EVE in that 30XAh range? Would looking up data sheets offer much of use to an aspiring but inexperienced enthusiast?

 

[zcskywire2]

Are there known differences of relvance between CATL, CALB, and EVE in that 30XAh range? Would looking up data sheets offer much of use to an aspiring but inexperienced enthusiast?

I or others have posted pretty much all of the datasheets for cells this size in the resources section of the site if you are interested. The noticeable differences are the cycle life and discharge/charge C rating. Some are 9000 cycles to 80% at 0.5c some are 4000 to 80% at 1C. The other notable thing to me is terminal shape, vent shape and QR code location. This allows you to tell if listings are actually showing pictures of what they claim to be selling. Also helps to know if a cell in a listing exists. There is no CALB 305Ah cell for example, those are just L173F280s being listed at their maximum test capacities.

 

[Coydon]

So I saw the Ganfeng 280s come back in stock at 18650batteries, marked as 270s and $95 each. 8 of these in a 24V configuration will put me around 6.5kWh, more than I think I'll need, while helping out the budget more than the 30Xs would. Sounds like they will be here any day.

I'm looking at the Overkill 24V 8s BMS to manage them. This seems to be pretty much the 'official' BMS of these forums?

Looking through the manual for Overkill BMSs, I noticed that the 24V does not include an internal temperature monitor like the 12V and 48V ones do. Anyone have an idea what that is about? I've been using LFP starter batteries in my motorcycle for years while doing some winter travel, many many times below freezing. I've talked with a manufacturer that has assured me the internal temps will rise to and stay at safe levels for charging as soon as it is started (though from dead cold running the headlight for a minute or two is recommended), even faced with an effective -gazillion wind chill temp in the ambient environment. Seems like if the BMS only measures externally, it will forgo charging in many situations where it should be safe. Does the limited discharge rate of these prismatics limit their ability to heat up internally?