LiFePo4 batteries two cells underforming all the time

Getting into the nuts and bolts of wind and solar power components, installation and performance issues, ask technical questions, answers and tips

LiFePo4 batteries two cells underforming all the time

Postby LRAT » Sun Mar 25, 2018 10:03 am

Dear enthusiasts,

I've got an array of LiFePo4 batteries: Two parallel banks of 16 cells in series. Total output voltage about 55V
Each cell is BMS monitored. Since a long time two cells have been under-performing compared to the rest. After a couple of days these two cells show a much lower voltage then the rest and then it trips the BMS.
Take for example yesterday: all cells were hovering around 3.20 V but the two weak cells were at 2.68 V and this caused the system to trip.
As an intermediate remedial I do the following:
I attach a regulated power supply to each under-performing cell
Set the maximum voltage at 3.5 Volts and this way I'm able to pull up the voltage in these two cells.

Now I've been thinking of building a device that can automate this procedure:
An Arduino board reads the voltage of a good neighboring cell and compares both voltages. If the voltage of the weak cell is lower then the good cell then a capacitor will be charged taking energy from the good cell. When the capacitor is charged the Arduino will then switch to the bad cell and discharge the capacitor to bump up the voltage. It will then compare the voltage of both cells again and will repeat the procedure until both voltages are the same.

Here's some logic:

read V1 and V2
if V2 < V1 then
charge capacitor
if capacitor voltage = V1
then disconnect from V1
Connect to V2
Discharge in V2
return
if V1 = V2 then do nothing

Do you guys think this is a good solution? Have you done something similar before? All tips and/or advice are welcome! Thank you very much in advance.
Cheers,
Gerard
LRAT
Solar Supporter
Solar Supporter
 
Posts: 3
Joined: Sun Mar 25, 2018 9:49 am

Re: LiFePo4 batteries two cells underforming all the time

Postby Gordon-Loomberah » Sun Mar 25, 2018 2:46 pm

Welcome to the Energy Matters Forums Gerard :)

What brand and capacity are your cells?

How long has your LiFePO4 battery been in use, and what sort of disharge do they typically get each day?

Are your parallel banks of cells tied only at the ends, or at each cell, or something in between?

When your cells were mostly at 3.2V, what was the SOC?

What BMS are you using? There are cell balancers that continuously move energy between cells, in order to keep the voltages close (ish) to each other, such as the one I am using, so no need to reinvent such a device.
http://gunagulla.com Loomberah weather and astronomy including live solar radiation intensity and UV + Gunagulla aquaponics, organic eggs and cherries
User avatar
Gordon-Loomberah
Community Moderator
 
Posts: 5762
Joined: Thu Sep 24, 2009 9:41 pm
Location: Loomberah NSW Australia

Re: LiFePo4 batteries two cells underforming all the time

Postby Warpspeed » Sun Mar 25, 2018 4:09 pm

Hello Gerard, welcome to the Forum.

I tested a "capacitor balancer" at one stage, and found that the energy transferred is not large, and the high peak currents through relay contacts are problematic.
Its fine to keep healthy cells in balance that are already in balance, but if cells have significant capacity difference, its not going to fix the problem.

The conventional shunt discharger is a much more effective method, but the voltage thresholds need to be chosen and set very accurately to get a very precise balance.
Warpspeed
Solar Crusader
Solar Crusader
 
Posts: 435
Joined: Fri Jan 03, 2014 3:18 pm

Re: LiFePo4 batteries two cells underforming all the time

Postby LRAT » Sun Mar 25, 2018 4:55 pm

Hi Warpspeed and Gordon,

I wouldn't use a relay as the switching speed will be too fast. I was thinking of using a MOSFET for the switching.
It takes a couple of days before the voltage drop in these cells becomes significant. So, if I can keep them inline with the others then it prevent this to happen.
To answer the questions of the other reply I received:
The batteries are a generic brand; I've got no idea what the brand is; there's nothing written on them.
The batteries have been in use for two years now and the capacity is 25 kWh.
The parallel banks are tied off at the ends.
At 3.20 Volts the SOC is around 85%
Thanks for your replies!
LRAT
Solar Supporter
Solar Supporter
 
Posts: 3
Joined: Sun Mar 25, 2018 9:49 am

Re: LiFePo4 batteries two cells underforming all the time

Postby Privatteer » Sun Mar 25, 2018 5:48 pm

Try moving the weak cells to either cell 16 or cell 1. Up to you which you think best. I personally chose cell 1 for the one I had that used to lag a little in charge and it sorted it out.

Cell balancers by EV power seem to do a decent job. I have my own picaxe unit on the fault output not that its ever tripped except under test.
Stage 1:
Fronius Primo 5 + 6.6kW Canadian Solar
Privatteer
Solar Crusader
Solar Crusader
 
Posts: 476
Joined: Sat Nov 19, 2011 12:00 am
Location: South West WA

Re: LiFePo4 batteries two cells underforming all the time

Postby LRAT » Sun Mar 25, 2018 6:01 pm

Hi Privatteer,

Thanks for that! The cell balancers you recommend are exactly the ones I got!

It will be a big job to separate the battery packs. Could you please explain the logic behind it before I attempt this job?
I thought, as they are in series, that charging would be simultaneously in all batteries.
Cheers,
Gerard
LRAT
Solar Supporter
Solar Supporter
 
Posts: 3
Joined: Sun Mar 25, 2018 9:49 am

Re: LiFePo4 batteries two cells underforming all the time

Postby Warpspeed » Sun Mar 25, 2018 6:07 pm

I use optically isolated mosfets (solid state switches) these have both good and bad points.

They are not all that fast, and need time to turn both fully on and fully off. There is also an "on" resistance of a few ohms which is not a problem when monitoring individual cell voltages, but it hugely limits the rate of individual cell charging and discharging to something the mosfet switches can handle.

Try feeding an amp through several ohms of series resistance, and you can see the problem.

Also, you need to think about possible failure modes where a software crash can randomly turn these mosfets on and off, possibly placing shorts across high voltage strings of cells. If only one mosfet device fails shorted, it can cause a cascading failure of every other mosfet device.

Mosfet opto switches CAN be protected with a 100mA fast blow fuse, I have blown a lot of fuses during development and never a mosfet switch. But a 100mA fast blow fuse is only useful; for voltage monitoring, and is useless for any serious energy transfer.

Not wishing to be a wet blanket, but I had to learn all this myself the hard way, and just wish to save you the same pain.

It's why I only use these mosfet switches for scanning cell voltages. Cell balancing is done with an individual shunt regulator bolted to the top of each and every cell.
Warpspeed
Solar Crusader
Solar Crusader
 
Posts: 435
Joined: Fri Jan 03, 2014 3:18 pm

Re: LiFePo4 batteries two cells underforming all the time

Postby Gordon-Loomberah » Sun Mar 25, 2018 6:59 pm

What colour are your cells? Any chance of a photo?

I had one CALB cell exhibit significantly reduced capacity a while back after about 3 years of use (details written up in another thread), and replaced it with a Winston, as CALB 400AH cells don't seem to be available in Australia any more.

If at 80% SOC they are only at 2.68V, then I would say they are stuffed, and it is not worth trying to revive them with any sort of BMS. That's a lot lower voltage than my weak cell was.
http://gunagulla.com Loomberah weather and astronomy including live solar radiation intensity and UV + Gunagulla aquaponics, organic eggs and cherries
User avatar
Gordon-Loomberah
Community Moderator
 
Posts: 5762
Joined: Thu Sep 24, 2009 9:41 pm
Location: Loomberah NSW Australia

Re: LiFePo4 batteries two cells underforming all the time

Postby Privatteer » Sun Mar 25, 2018 10:26 pm

Its just a personal theory I've developed over the years. Each battery has its own resistance which results in minor variations in cell voltages, you can measure this with the right meter. By placing it first in the string all current has to pass through the weaker cell giving it more chance to absorb. There is also the advantage as an outside cell it will generally be slightly cooler.
With shunts cell 16 might end up with 15 cells being shunted to it as an alternative idea.

I have played around with a "ICL7665S Super CMOS Micropower Over/Under Voltage Detector" chip as a settable balance unit. My initial prototypes did work for 4 cells but I never completed all 16.

I have to agree with Gordon however the cell voltages are indicating they are very much on the way out if they are dropping that far with the EVpower balancer's on.

What is your balance voltage and float voltage you are using?
Stage 1:
Fronius Primo 5 + 6.6kW Canadian Solar
Privatteer
Solar Crusader
Solar Crusader
 
Posts: 476
Joined: Sat Nov 19, 2011 12:00 am
Location: South West WA


Return to Components, Installation And Performance

Who is online

Users browsing this forum: No registered users and 6 guests

new solar power specials
cron