17s LiFePO4 battery

[szczepan]

Hi!

What do you think about a 17s LiFePO4 battery for a typical 48V (so 60V max in reality) inverter?

With a 17s setup charged to 60V, each individual cell will be around 3.5V, which is about 90% of its capacity, so it will never charge to 100%. Thus, I can utilize the battery in the 10-90% capacity range, which should be good for its longevity. In case one cell fails, I can easily rebuild it to a typical 16s, so this additional cell can be treated as some kind of "hot spare."
What do you think about this configuration?
A big drawback is the inability to charge an individual cell above 3.5V.

Has anyone tested such a setup?
Or is it better to simply keep one spare cell unconnected on the shelf?

 

[Sweet Tatorman]

3.5V is in fact 100%.

 

[BentleyJ]

It depends on the inverter and charge controller maximum DC voltages. For example Schneider XW is good up to 70V so theoretically an 18S battery would be possible. One of the difficulties with going over 16 cells is getting a single BMS with that many leads.

As to your proposal, charging to 3.5V will charge the cells to 100%. It may take a little longer but given sufficient time the cells will reach full charge.

 

[MisterSandals]

What do you think about a 17s LiFePO4 battery for a typical 48V (so 60V max in reality) inverter?

This game is mighty difficult doing things as recommended. What are you trying to accomplish by adding a 17th cell? Do you need more capacity or voltage?

Or is it better to simply keep one spare cell unconnected on the shelf?

Definitely! At least one.

 

[szczepan]

I try to accomplish:
- having spare cell
- using battery in 10-90% (or similar) range of capacity for it's longevity.

There are 17s and 24s DALY BMS:

Wholesale Daly Balance BMS 4S-24S 40A-500A For Lithium ion Battery Pack Manufacturer and Supplier | Daly

www.dalybms.com

I have also Deye inverter (currently connected to old Ni-Cd battery pack), so theoreticaly i can try to build such setup.

 

[BentleyJ]

Sorry, should have qualified my statement: "One of the difficulties with going over 16 cells is getting a QUALITY single BMS with that many leads".

 

[szczepan]

Sorry, should have qualified my statement: "One of the difficulties with going over 16 cells is getting a QUALITY single BMS with that many leads".

DALY isn't good BMS ?

 

[crezno]

I have tested this with a 20S and 24S JKBMS and it worked with both Li-Ion and LifePO4. I would say that whatever BMS you get, make sure it active balances.

Devils advocate, I have never had a set of cells degrade enough from full charge, full discharge, cold discharging, holding at full charge for weeks on end, that I needed to replace the cells. Seems like every 3 or 4 years, I end up getting the better cells anyway.

My oldest 2 battery packs are Renogy 100Ah bought in 2019 and they are at 97Ah and 95Ah respectively. (Back in 2019 I didnt have a way to get an accurate baseline of the full capacity so there could be 3% or 4% more degredation if they were actually 105AH capacity for example.) These have been cycling down to 30% up to 100% with solar every day for the past 3 years, and the prior 3 years they were in a camper sitting at full charge minus the 15-20 days it was cycled when it was being used.

 

[mikefitz]

around 3.5V, which is about 90% of its capacity, so it will never charge to 100%.

If the cell reaches 3.5 volts, its at 100% charge, not 90%.

 

[crezno]

Could try to do a 17S and set charge cutoff voltage at 57v, then resting voltage may come back down a bit and be in the 85% range +/- who knows because this is hard to really know without knowing actual full capacity and using a coulometer and the environment being stable every day.

 

[koppted]

Hi!

What do you think about a 17s LiFePO4 battery for a typical 48V (so 60V max in reality) inverter?

With a 17s setup charged to 60V, each individual cell will be around 3.5V, which is about 90% of its capacity, so it will never charge to 100%. Thus, I can utilize the battery in the 10-90% capacity range, which should be good for its longevity. In case one cell fails, I can easily rebuild it to a typical 16s, so this additional cell can be treated as some kind of "hot spare."
What do you think about this configuration?
A big drawback is the inability to charge an individual cell above 3.5V.

Has anyone tested such a setup?
Or is it better to simply keep one spare cell unconnected on the shelf?

The Midnite One, SungoldPower: SGN11.4KHB-48, the Midnite Rosie, Can charge up to 64v+ and Solark 15K can Charge up to 63v. For 17s. If you want to top balance to get voltage up to 3.65 per cell or 62v for the pack. Any one of those inverters will work. Edit I should mention the discontinued XWPros. Can also do the same.

​

 

[frankrudis65]

Hi there,I've got 17S2P since just over a year.There's no draw back,even running them at lower voltage.No balancing issues either I've seen a guy on youtube,who kept their lifepo4 voltages between about 3.42,with a voltage drop of 0.5 for float.This was successfully done for 17 years.I was haply do adapt ,as this long-time test convinced me. As my Victron has got no problems with a possible 60V charge,a 3.5V per cell wouldn't be a problem .As I don't need this,I keep the cells at about 3.425V.Appart from having things running "cooler" especially in the summer months ,and a little extra possible energy,if one cell would go bad,it's easily put to 16S..JK bms (not the inverter one) has got no problems with 17S.Having a decent load of 4.5Kwh on the battery banks doesn't bother them in the slightest .Best regards Frank

 

[LydMekk]

DALY isn't good BMS ?

No.

 

[szczepan]

Hi there,I've got 17S2P since just over a year....

Thanks for your reply.
I've ordered 17 x 105 Ah cells and Dally 100A BMS.
Will see how it performs with Deye and Easun inverters.

 

[Ikassir]

Will see how it performs with Deye and Easun inverters.

we are looking for hearing about that
good luck

 

[SolitaryPhoton]

I might give 18s a try with my EG4 3000 AIO. I have some spare cells lying around and my BMS can handle it. The settings for the inverter are just high enough too.
Max voltage cuttoff: 54V (3V cell)
Max bulk voltage: 62V (3.45V cell)
The equalization setting can go up to 64V (3.55V cell) if balancing is an issue. As long as I don't blow up my inverter...

 

[szczepan]

I might give 18s a try with my EG4 3000 AIO. I have some spare cells lying around and my BMS can handle it. The settings for the inverter are just high enough too.
Max voltage cuttoff: 54V (3V cell)
Max bulk voltage: 62V (3.45V cell)
The equalization setting can go up to 64V (3.55V cell) if balancing is an issue. As long as I don't blow up my inverter...

I have Easun inverter which has 62V max but my second inverter is Deye maxed at 60V - so i decided to make 17s.

 

[SolitaryPhoton]

Just reporting back. Got the two extra cells in and everything seems to work. Inverter didn't blow up or anything

 

[frankrudis65]

Hot nights in front of You

 

[Ikassir]

17s LiFePO4 battery for a typical 48V

inverters doesn't support higher battery voltages because there is no batteries with this range , not because it is difficult from the technical POV
so the reason why usually we don't go beyond 60 because it start to be high voltage battery . the threshold for DC electrical shock for dry skin is 50-60 V , so if we are ok with more danger batteries why not go high voltage inverter and batteries

 

[1ufvlsi6x]

17s is ludicrously incompatible with the best DIY products right now in terms of cost/quality. For instance the best cost/quality BMS for DIY is right now the 16S JK inverter series and it won't work. There is an entire industry around 16s in general that it won't work well with 17s so I wouldn't even START.

 

[1ufvlsi6x]

@42OhmsPA well if you're unsure, do it, you'll be sure later.

 

[42OhmsPA]

@420Heysus well if you're unsure, do it, you'll be sure later.

I'm betting you tagged the wrong person, I'll reply.

I've already considered it many times, if I bought enough spare cells I would have already done it, it makes great sense. Unfortunately I don't have enough spare cells to add to all my packs and I build my racks to handle 16 cells wide...

Multiple BMS are available that can be wired up to 21S, the JBD AP21S that I use for example. The OG JK BMS can go up to 24S.

Multiple AIOs will charge up to 60V, or more.

3.5V × 16 cells = 56V for a fully charged pack.

 

[DIYrich]

On the sol-ark 15k, the stated limit off battery is 12kW, but that assumes 48v at 250a from the battery. The 15k can pull 52v at 275a, or around 14kW. If you can boost to 55v (17s), you can get the full 15kW off battery only for most of the discharge range of the lifepo4 battery. 18s could also work, but you only charge to 3.33v/cell at 60v.

 

[1ufvlsi6x]

Multiple BMS are available that can be wired up to 21S

I didn't say it can't be done correctly, in isolation in 1 system. I said it's dysfunctional for modularity in the future and having the best options in the market.