diy solar

diy solar

Building a 72KwH battery system, help understanding these Li-ion modules.

hummmingbear

New Member
Joined
Feb 8, 2022
Messages
172
Quick background: I live off-grid and am ready to replace my 12-year old Flooded Lead Acid batteries that our home/systems run on. I'm looking at buying 35 of these Samsung EV modules from BatteryHookup which would give me 72KwH of storage. They are offering me a bulk discount and free shipping which is making it more enticing. I understand my FLA batteries quite well, but I don't have the same depth of knowledge on Li-ion batteries to make an educated decision here, so I'm hoping folks here can give me feedback.

I'm not quite sure what cells they are used in these batteries, does anyone know? Does Samsung make good Li-ion cells that would be good for off-grid home use? At 35 modules I shouldn't have to work about pulling too many amps from these? My inverter is 5000VA but I rarely peak over 3000W.

These modules are 48V, but I asked and you can "cut the bus bar between the 7th and 8th cell" to make them 24V. My inverter is 24V and I would prefer not to replace it. If the only benefit to 48V is smaller wiring, fusing and BMS it will still be more affordable to pay extra for those things than to replace my inverter. Is there anything else that I'm not seeing that makes it a bad idea to make these modules 24V?

Anything else I might be missing? I'm aware of the basic for Lithium Ion...keeping them above freezing, risks of thermal runaway if they get to hot. What else should I know before I make the investment into these? Any help appreciated.
 
250 * 35 = $8750 - not counting BMS + shipping (was $677 for me)

Due to the nature of that chemistry, if you want extended cycle life, you're only you're going to be able to utilize 50-60% of the capacity. Non-LFP Lithium chemistry does not have the same cycle life and demands short cycling in the 20-80% SoC range for comparable longevity to LFP.

So, for 72kWh of storage, you'll get 43.2kWh usable

For 43.2kWh of usable LFP, you'll need about 54kWh of LFP

NEW 25.6V LFP in 280Ah cells (7.2kWh) will cost you about $1300 shipped including BMS (you'd have 8 separate 25.6V batteries each with their own BMS)

So, you can buy 57.3kWh of NEW LFP with 46kWh usable for $10,400 vs. USED non-LFP Lithium for $8750 (+$677 shipping) with HOPEFULLY 43.2kWh usable.

LFP: $226.09/kWh (usable)
USED non-LFP: $218.21/kWh (usable) + BMS cost + additional labor.

Perspective: I'm someone struggling to build a Lithium NMC-LMO battery (from PHEV) due to the labor involved (588 cells). I spent about as much as I would have on the equivalent amount of LFP, but I would have been done a year ago.
 
250 * 35 = $8750 - not counting BMS + shipping (was $677 for me)

Due to the nature of that chemistry, if you want extended cycle life, you're only you're going to be able to utilize 50-60% of the capacity. Non-LFP Lithium chemistry does not have the same cycle life and demands short cycling in the 20-80% SoC range for comparable longevity to LFP.

So, for 72kWh of storage, you'll get 43.2kWh usable

For 43.2kWh of usable LFP, you'll need about 54kWh of LFP

NEW 25.6V LFP in 280Ah cells (7.2kWh) will cost you about $1300 shipped including BMS (you'd have 8 separate 25.6V batteries each with their own BMS)

So, you can buy 57.3kWh of NEW LFP with 46kWh usable for $10,400 vs. USED non-LFP Lithium for $8750 (+$677 shipping) with HOPEFULLY 43.2kWh usable.

LFP: $226.09/kWh (usable)
USED non-LFP: $218.21/kWh (usable) + BMS cost + additional labor.

Perspective: I'm someone struggling to build a Lithium NMC-LMO battery (from PHEV) due to the labor involved (588 cells). I spent about as much as I would have on the equivalent amount of LFP, but I would have been done a year ago.
FWIW I can get the 72Kwh of these cells shipped for $7000. I still need a BMS. That is still quite a bit cheaper than the $10,400 - With that price do you still feel similar?
 
I still need a BMS.
Nope, you need 35 BMS's, one for each module, unless you can get all the cells from each module in parallel, but it seems that they're welded in series, so that will make you spend more time and money. Those modules will be good for a high voltage application, like over 300V battery pack, for electric vehicles.
 
If you want to run just one BMS, you have to parallel groups of 70 cells first, and then put those groups in series, 70P8S, for 24V.
 
Now you need 70 BMS's :unsure:
Ohhhhhkay. Yeah this is the information I needed. Doesn't seem wise to have 70BMS's!

If you want to run just one BMS, you have to parallel groups of 70 cells first, and then put those groups in series, 70P8S, for 24V.
This would require pulling and re-doing the existing spot welds? If so that's exactly what I was hoping to avoid...

Those modules will be good for a high voltage application, like over 300V battery pack
Can you explain why that is? What specifically makes these geared for high-voltage? If my inverter supported ~40-50v wouldn't I just connect these all in parallel as they are? Appreciate your help, thank you.
 
FWIW I can get the 72Kwh of these cells shipped for $7000. I still need a BMS. That is still quite a bit cheaper than the $10,400 - With that price do you still feel similar?

Absolutely.

These are made from 18650 cells, likely NCA chemistry - the most dangerous and shortest full discharge cycle life cells available. Their only real advantage is power density, i.e., they're lighter per kWh.

Why not getting the 48V one? https://batteryhookup.com/products/...98914&pr_ref_pid=7324360769698&pr_seq=uniform

And be aware that they're one of the chemistries that can catch on fire is something goes wrong.

Because 16S on this chemistry is typically unusable for most inverters if one ever wants to take the cells to full charge.

Can you explain why that is? What specifically makes these geared for high-voltage?

They come out of EVs where modules are stacked in series to attain 300V+ with a centralized management system

If my inverter supported ~40-50v wouldn't I just connect these all in parallel as they are?

Yes, but you'd still need 35 BMSs.
 
I just would not trust those Samsung cells and that vendor has been having issues.
If you are willing to roll your own, look at the Eve 230Ah cells and JBD BMS from Docan Power - https://www.docanpower.com/index.php?route=product/product&product_id=434
These cells are the best bang for your buck and in stock in the USA. They can be had cheaper shipped from China from the same supplier.
The BMS is slightly daunting as the documentation is poor but we have it figured out. It is a very good BMS - https://www.docanpower.com/index.php?route=product/product&product_id=444
You can put 2 cells in parallel as 8s2p for a per battery of 11kWh @ 24v and I have built this for $2060ish but in the 48v format.
As others have said, you will have much more usable capacity, longer life and a safe non venting battery, so 55kWh of battery for about $11k will out perform the AGM long term.
 
Ok...thank you all for your advice here. Clearly these do not seem like the right application for my situation.

While I have you here...I have been eyeing the Eve LiFePO4 cells. I'm sure you're familiar with them around this forum. If my math is right I would need put them in sets of 8S to make them 25.6V, and then parallel my 8S, my budget affords about 56 of these modules giving me 8S7P. Can I use 1-BMS to manage all of these? I'm sticking with ~24V to avoid replacing my inverter.
 
Ok...thank you all for your advice here. Clearly these do not seem like the right application for my situation.

While I have you here...I have been eyeing the Eve LiFePO4 cells. I'm sure you're familiar with them around this forum. If my math is right I would need put them in sets of 8S to make them 25.6V, and then parallel my 8S, my budget affords about 56 of these modules giving me 8S7P. Can I use 1-BMS to manage all of these? I'm sticking with ~24V to avoid replacing my inverter.

Those are the cells I've been talking about. Note how 8 cells + a BMS is about $1300 :)

Sorta, but it's not recommended. You'd need to parallel sets of 8 and then series each group. Your system would be limited to about 5000W maximum current, and you would have NO redundancy.

I would likely consider 2P8S X 4
 
So with my suggestion and your budget you can build 44kWh of battery complete.
You can always add more later if required.
You can always add more later comes with an asterisks yes? Isn't the limit to adding new cells to an existing group about around 2-years? Thank you all for the help, really appreciate this forum.
 
Back
Top