No BMS on LiFePO4 cells

[flatfootskier]

Hi,
I'm looking to build a grid tied storage system for home: PV + Inverters + Battery - 48V from 16 x LiFePO4 @ 200Ah.
When I get the batteries I intend to:

1. Measure individual cell voltage & Internal impedance. I will closely match the cells & use the leftovers for other things.
2. Put them all in parallel for an initial charge & balance
3. Series them up & start using them - but I'm contemplating without a BMS - Reason: I don't trust the cheaper ones & can't justify the top end ones.
4. I believe that by keeping the batteries within 15-85% of capacity I should avoid the position where one cell reaches either empty or full & starts getting damaged as its neighbouring cells continue charging/discharging
5. I intend putting per-cell monitoring of voltage & temperature in place and having external alarm & control (probably Arduino).
6. At a later point I'll consider a relay & large resistor across each cell, if required, for balancing.

Comments welcome. I'm trying to demistify some of the nonsense floating around. Thanks in advance,
Simon

 

[Diver300]

If you've got a monitoring system to make sure that no cell goes below 15% or above 85%, you've got a BMS of sorts.

The discharge curve of LiFePO4 is fairly flat until fully charged or fully discharged, so you won't know if the cells are out of balance until one gets fully charged or fully discharged.

 

[rjenkinsgb]

See the Battery University info on LiFePO4:
**broken link removed**

Extract from that:

With most batteries, cold temperature reduces performance and elevated storage temperature shortens the service life, and Li-phosphate is no exception.

Li-phosphate has a higher self-discharge than other Li-ion batteries, which can cause balancing issues with aging.

The battery pack cost appears to be several thousand pounds!

Only monitoring the overall series voltage as you appear to be suggesting, can not guarantee the individual cell voltages.

Trying to save a relatively small percentage of the battery cost by not having the proper management setup seems false economy; you want to be aiming for the maximum possible battery life and reliability.
Add a BMS from the start; It will save money in the long run.

 

[flatfootskier]

The battery pack cost appears to be several thousand pounds!

Only monitoring the overall series voltage as you appear to be suggesting, can not guarantee the individual cell voltages.

Trying to save a relatively small percentage of the battery cost by not having the proper management setup seems false economy; you want to be aiming for the maximum possible battery life and reliability.
Add a BMS from the start; It will save money in the long run.

Thanks for the link, hadn't seen that.

I intend relying on overall series voltage for control decisions re charging & discharging, but you might not have spotted this:

5. I intend putting per-cell monitoring of voltage & temperature in place and having external alarm & control (probably Arduino).
6. At a later point I'll consider a relay & large resistor across each cell, if required, for balancing.

I understand it's over £1000 for a Batrium, which seems to be the Rolls Royce BMS at the standard I would like. That's not a small percentage of my build. I don't want to trust the whole setup on some crappy Chinese BMS where I don't know it works or what its safe limits are.. My peak current is 280A. Hence my intention to run within 15-85%, where, as I understand it, a BMS wouldn't do anything anyway. I'll be watching voltages like a hawk for the first few weeks! And I plan to introduce an alert if any cells fall out of alignment.

I have moderate electrical/electronics knowledge - enough to know that there's lots of sketchy stuff going on in this field. I'm keen to validate whether remaining within the 15% to 85% window is, in the short term, a viable first principle strategy, whilst I learn from the monitoring & design something fit for purpose without remortgaging!

 

[Pommie]

What size battery bank will you use? Voltage AH etc.

Mike.

 

[flatfootskier]

What size battery bank will you use? Voltage AH etc.

Mike.

48V from 16 x LiFePO4 @ 200Ah.

 

[Pommie]

Is that 16 x 200AH cells so about 12,800 WH?

Mike.

 

[alec_t]

Hence my intention to run within 15-85%

How will you know the state of charge? Coulomb counter?

 

[Pommie]

Here, in Australia, most LiPo batteries are the 12V type with built in BMS and so not suitable for solar etc - unless you run at 12V. I'm curious what the OP has.

Mike.

 

[flatfootskier]

Is that 16 x 200AH cells so about 12,800 WH?

Mike.

correct

 

[flatfootskier]

Here, in Australia, most LiPo batteries are the 12V type with built in BMS and so not suitable for solar etc - unless you run at 12V. I'm curious what the OP has.

Mike.

LiFePo4, not Lipo. You can get this chemistry in a complete *battery* but I'm getting *cells*

 

[flatfootskier]

How will you know the state of charge? Coulomb counter?

I'll have coulomb counting, but only for capacity indication. Charge control will be done by a victron MPPT controller, which allows configurable settings, based on Voltage, much like lead acid. I don't plan on getting very close to the 3.65V maximum voltage.

 

[CG1]

Maybe you could use just one big cell (100000mAh) and a DC-DC boost converter. No BMS needed.

 

[Pommie]

Maybe you could use just one big cell (100000mAh) and a DC-DC boost converter. No BMS needed.

A 100Ah single cell will store ~400Wh which is much less than the 12kWh the OP indicated.

Mike.