Aging LiFePo4 Question

I've seen a few discussion that have had the occasional indirect reference to the gist of this question, but no definitive agreement.

As LiFePo4 batteries age, and basically just about any other battery, they lose capacity. My question for LiFePo4 batteries is how this capacity loss is reflected in the battery characteristics?
Is there simply a loss of amp hour capacity with no impact on voltage profile? Or does the maximum voltage start to drop as the battery ages? If it is the latter then it would seem that over time one would have to modify the charging configuration on the SCC to accommodate this.

I have two used BYD batteries and I want to ensure I use the correct charging configuration for these batteries. My intent was to keep them in 10-90% range in order to get the maximum number of cycles out of them.

I'd appreciate comments from those in the know or who have had experience with this.

Hi, I have 4 BYD packs, this is what happens.

Below is a chart of the individual cell voltages as the aggregate voltage
limit is reached. You can see that the most worn cells start rapidly increasing their voltages, much faster than the healthy cells.

At the graph midpoint, charging is stopped and the voltages converge, then at the end, a load is briefly applied before the charging starts again.

This cell voltage divergence in effect shuts down your battery bank even though the healthy cells could accept more charge.

Balancing the cells can help, but not as much as you think. Your battery is like a rusty chain, its as only as strong as the weakest link. When one cell goes bad, it brings down the whole pack.

The trick is to monitor the cell voltages like I am below with a Chargery BMS and stop charging when the cell voltages
diverge like below (ie > 100mv), this limits further cell damage.

So as you commented, the actual effective peak voltage drops over time. Im around 70% capacity on my BYDs and
can get 3.8kw max out of each one. Have you capacity tested yours?



This is my daily charging report for my BYD pack, my software shuts down charging when the cell divergence gets out of limits.

That has been my experience with Nissan Leaf modules as well. The weaker modules limit the pack capacity because you have to lower your CV (Absorb) set point to avoid those modules going above a safe voltage. I rarely get near the bottom but the same thing happens there.

I think we are in uncharted waters with these BYD packs. I bet these packs were cycled a high current, with the massive heat syncs and fans. The negative buss bars look to be discolored from heat.
I am getting 100ah out of my packs without any bms, the lower capacity cells are testing around 110ah and the others range between 110-140ah.

I have settled into daily charging to 54.4v absorbtion for 5min, and float all day at 53.8v. At end of day the cells are all within 20-60mv with no bms and trimetric reads 100%.
Hard to say charging a 50% degraded cell to 100% of usable capacity is hard on the cell or not. My best guess is as long as the cell are not going way out of balance, and not charging at a high rate they could last for a very long time.. If they were going to just fade away I think they would have done it by now.. But as you can see there is a lot of.. I think.. Guess.. Who knows.. In my reply..

Mtour said:

I think we are in uncharted waters with these BYD packs. I bet these packs were cycled a high current, with the massive heat syncs and fans. The negative buss bars look to be discolored from heat.
I am getting 100ah out of my packs without any bms, the lower capacity cells are testing around 110ah and the others range between 110-140ah.

I have settled into daily charging to 54.4v absorbtion for 5min, and float all day at 53.8v. At end of day the cells are all within 20-60mv with no bms and trimetric reads 100%.
Hard to say charging a 50% degraded cell to 100% of usable capacity is hard on the cell or not. My best guess is as long as the cell are not going way out of balance, and not charging at a high rate they could last for a very long time.. If they were going to just fade away I think they would have done it by now.. But as you can see there is a lot of.. I think.. Guess.. Who knows.. In my reply..

Click to expand...

You really are gentle on those cells!! Do they charge at all at 3.36?

zorlig said:

You really are gentle on those cells!! Do they charge at all at 3.36?

Click to expand...

I am going to be charging to 3.4 volts with new LFP cells. What is your theory that supposes they would not accept some current at 3.36 ?

Ampster said:

I am going to be charging to 3.4 volts with new LFP cells. What is your theory that supposes they would not accept some current at 3.36 ?

Click to expand...

I don't, I'm just curious if they do continue to charge after you drop from 3.4 and if the batteries eventually reach a full charge in your testing. If there is a voltage below 3.4 where they still fill everything then I'd love to know what that is

I would love to see charge curves on new cells at various float voltages! I'll have to make them some day

3.36v looks like a .2C at 30% .1C at 50%, .05C rate at 70% fill.

What C rate are you running the 3.4% at for 5 minutes? Yours seems like a decent plan to keep the battery from exceeding a reasonable state of charge as long as there is at least some daily cycling!

I am charging without a bms, it would be great to have a qnbbm on each pack but at $200.00 per it is not an option. Maybe one day there will be an active balancer that is reasonably priced..

I have done tests on capacity of single cells and found that from 3.4v to 3.6v it only takes between 8-10amps. It's just not worth the worry to eek out every amp, especially in the summer months. This will probably change to a more aggressive charge in the winter months,

I am happy with 100ah per pack capacity, the max charge on solar is 20ah per pack. My Batts are cycling between 2.9v-3.4v, mid voltage all cells are within 40mv and out as much as 60-80mv on the top and bottom. I want this to be a set it and for get it system as much as possible..

I am monitoring with a Bogart trimetric, which I would not go without. It's reading 53.9v..98% and .8 amp charging.

Well, I just finished cycling my pair of BYD's and here are the results. Bear in mind I'm running a 24V system and the corresponding cells in the batteries have been hooked in parallel, so it behaves like a 2P8S.
I bottom balanced at 24 Volts, using my BattGo-8S. Yeah I know, I got it more to see my cell voltages. Took 5 days to get balanced.
I've cycled through a discharge/charge cycle twice now and here is the problem. Two cell pairs are hitting 4 volts around the time the battery shows 28 volts. As a result I'm only getting a bit over 200 Amp Hours, or 4.8 KwH out of these two batteries each originally rated at 5.6 KwH.
This is without a BMS, other than the BattGo, which effectively isn't doing any balancing.
I stop charging at that point and after the battery settles over a few hours the voltage drops to 27.9 volts. Any attempt to restart the charging drives the two weak cells ver 4 volts in a few minutes.
I've read the long BMS discussion and concluded a few things. I can't use the cutoff from a BMS because I'm using an All-In-One soltution so the SCC and inverter share the same input leads. No way to cut off charging without cutting off the inverter.
Still, if I can find an aggressive active balancing BMS I can likely charge these batteries further, while maintaining a safe charge profile on the SCC, but not using the BMS cutoff. Anyone know which BMS balances the most aggressively?

Qnbbm will do 6amps, or add a couple of headways to the weak cells. Qnbbm is the strongest active balancer I have found, if you can stomach the price of it. I got my uses headways from bch, they tested 100% capacity..

Unfortunately bch and bhu both are sold out of the headway’s at this time. I’ll take a look at qnbbm.

Just set your charging voltage lower and live with the lower capacity. It is a used pack. You can't restore capacity in a weak cell group.
All that is being done with an active balancer is that the current to that weak cell is being retarded while the other cells catch up. The end result of that process is that you will have a top balanced pack after all that effort of bottom balancing. No capacity will be added because the weak cell will now hit the bottom first and your Low Voltage Diconnect will have to be moved up.
Think of a group of sticks in your hand that have different lengths. When you bottom balanced them it was like tapping the sticks until they were even at the bottom and the top was irregular. Turn tour hand over and make the top level. Now the bottom will be irregular.

It is my understanding that top and bottom balancing are more useful when a BMS is not being used. The BMS has the effect you mentioned of top balancing at the top of charge, but also has the effect of bottom balancing at the bottom of charge.
Assuming the active balancer is sufficient to maintain a reasonable balance with the expected charge and discharge rates then it should be possible to run close to top and bottom SOC.
While charging the balancer moves the power from the fast charging (lower capacity) cells to the slower charging (higher capacity) cells, thus allowing them to store more power without overcharging the weak cells.
While discharging the balancer is transferring power from the slower discharging (higher capacity) cells to the faster discharging (lower capacity) cells, thus allowing one to draw more power from the battery without over-discharging any cells.
So, if I have two cells with only 80% of the capacity of the other cells, without a balancer I get 80% out of all the cells, and thus only 80% of the total new capacity of the battery.
With a balancer, I get 80% out of the 2 weak cells, but 100% out of the stronger cells, and thus 95% of the total new capacity of the battery.
Now, that said I will still not run the full range of the SOC, but only 10-90%. My current issue is that I am currently not able to get anywhere the original new capacity, but I seem to only have 2 bad cells.
What I am achieving is closer to 5.1 kWh out of what was originally a total of 11.2 kWh of batteries. So, I'm willing to go a little further to get a bit more out of these. A qnbbm 8S is about $200. I expect that I would get at least another kWh out of these batteries, given the SOC of the other cells. $200 for a kWh is cheap given todays prices.
Thank you to all for your feedback. It is greatly appreciated.

My BMS only works at at an upper voltage that I can set for it to start shunting current. I do not know of any way for it to do any form of shunting at the bottom. I guess that means it is a passive balance. I guess that you are looking for more of an active balancer. If you get that result it it is $200 a kWh but you are recovering value from your sunk costs.

I have a similar issue with some Nissan Leaf modules that I acquired for $150 per kWh. Fortunately I can just remove the weak ones from my pack but I actually need more capacity. I have committed to purchase a new 30 kWh pack for less than $125 per kWh and I think I can sell most of the Leaf modules for $120 per kWh.

You can't beat headways for the cost, at $1.00 per ah for lost capacity per pack. A couple of my packs needed two headways added on a single cell, at 8ah per cell the two gave 16ah to the one cell which in turn let me get 16ah out of every pack in the whole power wall for a $32.00 investment..

Mtour said:

You can't beat headways for the cost, at $1.00 per ah for lost capacity per pack. A couple of my packs needed two headways added on a single cell, at 8ah per cell the two gave 16ah to the one cell which in turn let me get 10kw out of the whole power wall for a $32.00 investment..

Click to expand...

I am not sure I understand the math? The Headways added 51 Watthours. (16Ahrs x 3.2 volts) and cost $32. My math says that is a cost of $627 per kWh for the Headways. How much was your pack capacity before you added the Headways? I also presume you mean you got 10kWhs out of that pack.
Don't misunderstand, I am a big fan of Headways, I built a lot of ebike packs with Headways and they lasted a long time.

Mtour said:

You can't beat headways for the cost, at $1.00 per ah for lost capacity per pack. A couple of my packs needed two headways added on a single cell, at 8ah per cell the two gave 16ah to the one cell which in turn let me get 16ah out of every pack in the whole power wall for a $32.00 investment..

Click to expand...

Now, if only the headway’s were available used.

Ampster said:

I am not sure I understand the math? The Headways added 51 Watthours. (16Ahrs x 3.2 volts) and cost $32. My math says that is a cost of $627 per kWh for the Headways. How much was your pack capacity before you added the Headways? I also presume you mean you got 10kWhs out of that pack.
Don't misunderstand, I am a big fan of Headways, I built a lot of ebike packs with Headways and they lasted a long time.

Click to expand...

I am running 192 cells..

Mtour said:

I am running 192 cells..

Click to expand...

As mtour is pointing, allowing the one weak cell to effectively accept 8 more amp hours, means all other strong cells can also accept that much more. So 192 cells times 8 amp hours.