diy solar

diy solar

When do you discard your LiFePO4 cells

SolarMinerPH

New Member
Joined
Sep 21, 2021
Messages
17
Some datasheet that I have seen says that after reaching its max cycle that it will have 80% of its original capacity. And then some articles I have read says that they replace their batteries when the total capacity of the battery is down at 80%.

My question is if it still has 80% of capacity then why do you need to replace it? It still has 80% capacity right? Can't you just parallel a new set of battery bank with its own BMS to offset the lost 20% capacity. So if your original battery bank is 100AH and you add another 100AH then now you have 180AH.

Is the internal resistance too high at that point that you can no longer discharge it at a decent rate?

I would think that you would only discard the batteries once it can no longer hold any charge.
 
Some datasheet that I have seen says that after reaching its max cycle that it will have 80% of its original capacity. And then some articles I have read says that they replace their batteries when the total capacity of the battery is down at 80%.

My question is if it still has 80% of capacity then why do you need to replace it? It still has 80% capacity right? Can't you just parallel a new set of battery bank with its own BMS to offset the lost 20% capacity. So if your original battery bank is 100AH and you add another 100AH then now you have 180AH.

Is the internal resistance too high at that point that you can no longer discharge it at a decent rate?

I would think that you would only discard the batteries once it can no longer hold any charge.
The used lithium cell market from EVs such as the Nissan Leaf is a case in point. At 80% original capacity you can purchase a new battery for your EV and exchange the old battery for credit.

Nissan sells those degraded battery cells on the used cell market and outfits like TechDirect sell batteries based on those used cells at a ‘capacity’ representing about 80% of what those cells could hold when they were new: https://www.techdirectclub.com/48v-lfp-120ah-6kwh-lifepo4/

So with age, the energy storage density of these cells including LiFePO4 has decreased but if you have the space/volume to add additional cells / batteries to compensate, their is no distinct ‘end’ to their useable lifespan…
 
Is there a point where thermal runaway is an increasing risk?
Not worried about losing capacity.
It’s a good question and I don’t have an answer for you, but as cells age/degrade, I believe that translates to a higher and higher % of free Lithium ions getting stuck/trapped (primarily within the carbon-bass anode).

With less free lithium to store and later discharge charge, I don’t see any reason the risk of thermal runaway should increase.

My uderstanding is that thermal runaway is more related to the formation of dendrites which are the result of abusing the battery.
 
With LiFePO4 I’m not aware of any thermal runaway issues.

My oldest LiFePO4 battery is about 7 years old and has minimal diminished capacity which does not affect my use at all.

I plan to run it until it can’t hold a charge.

Assuming I live that long ?.
 
It's not the 80% capacity loss that hurts. It is the ability to deliver moderate current. The cell voltage will collapse with greater loads,.

When a cell gets old the capacity number measured is very dependent on load current applied during capacity test. Should test cells at 0.2 to 0.4 CA current rate. Most low cost capacity testers only do 20 to 40 amps so for a 280 AH cell that is only 0.07 CA to 0.14 CA. 40 amps discharge is bare minimum to get a reasonable idea of a 280AH cell quality.
 
It's not the 80% capacity loss that hurts. It is the ability to deliver moderate current. The cell voltage will collapse with greater loads,.

When a cell gets old the capacity number measured is very dependent on load current applied during capacity test. Should test cells at 0.2 to 0.4 CA current rate. Most low cost capacity testers only do 20 to 40 amps so for a 280 AH cell that is only 0.07 CA to 0.14 CA. 40 amps discharge is bare minimum to get a reasonable idea of a 280AH cell quality.
Isn't the voltage drop during high current loads due to the cell's higher internal resistance? If it is, then an internal resistance tester should provide enough insights on the cells health. Or are there other factors that would cause the cell voltage to drop drastically under higher C rates?
 
Well, I am still running FLA at my off grid place. But the criteria for replacement has always been when putting the coffee pot on in the morning causes a low voltage disconnect. :p
 
What you see with a YR1035 battery impedance meter is mostly electrode material resistance (LFP & graphite with their binders) , metal foil, and terminal resistance with some percentage of cell ion diffusion resistance. Almost none of the ion creation energy (de-intercalation energy) is included because for AC current there is not much ion creation, only existing ion transfer diffusion energy to move ions back and forth with the AC current. The ion diffusion resistance gets worse as cell ages so 1 KHz impedance does give a trend of battery degradation but the reading of a YR1035 1KHz impedance meter only represents a portion of the terminal voltage slump.

The percent of total terminal voltage slump it represents depends on current demand and age of cell. In terms of IR voltage drop it represents from about a quarter to half of total terminal voltage slump under cell load current. As a cell ages the metal foil resistance does not change much (unless it delaminates from electrode) but the diffusion and ion creation energy goes up.

YR1035 battery impedance meter is good at cell quality evaluation but not terminal voltage slump prediction.

In the graph below, the YR1035 impedance reading would be a little above the purple initial T=0 line. As with initial IR drop, the AC impedance does not have a time decay delay like the final terminal voltage slump does. This graph represents a fairly new cell. A used cell may have 3 to 5 times the voltage slump for moderate load current. At greater then 5x voltage slump from new cell, it is ready for trash recycling bin as the cell will not support much load current.

LF280 overpotiential curve.png
 
Last edited:
Nissan sells those degraded battery cells on the used cell market
I was not aware of that. I bought mine from Craigslist and in one case the whole pack came out of a wrecked Leaf.
Mine had less than 80% capacity. I did hear some reports of faster deterioration due to dendrite growth but I don't know the specifics. Because of that I transitioned to LFP and in response to title of the thread, I would tend to use LFP chemistry longer than I would trust NMC.
 
I was not aware of that. I bought mine from Craigslist and in one case the whole pack came out of a wrecked Leaf.
Mine had less than 80% capacity. I did hear some reports of faster deterioration due to dendrite growth but I don't know the specifics. Because of that I transitioned to LFP and in response to title of the thread, I would tend to use LFP chemistry longer than I would trust NMC.
There are reports of accelerated battery degradation on Nisson due to lack of a good battery thermal management system. Quite a few owners are complaining of significantly reduced driving range after car is two years old.

If you can tolerate the PhD professor presenter, the Youtube video "Why do Li-ion Batteries die ? and how to improve the situation?" it is talked about near timestamp 25:50 into video. This video pretty much focuses on electrolyte impact, and the search for 'special sauce' additives to extend cycle life.
 
Last edited:
With LiFePO4 I’m not aware of any thermal runaway issues.

My oldest LiFePO4 battery is about 7 years old and has minimal diminished capacity which does not affect my use at all.

I plan to run it until it can’t hold a charge.

Assuming I live that long ?.
after so many years of easily outliving my phone batteries, LiFePO4 chemistry has me thinking in longer time scales too. really turns things on its head compared to what’s normal in most electronics.
 
if i understand RCinFLA correctly.

for a given cell/pack.

apply 0.2-0.4CA load to it and observe terminal voltage slump = Y millivolts. “start of life reference voltage slump”

later when cell/pack is older.

apply 0.2-0.4CA load and observe 5 * Y millivolts drop? good time to decommission “end of life voltage slump”

of course with bigger cells it might be something like 50A. but the important part is that the amps are same between reference and check and terminal voltage slump is 5 times what was observed at start of life

*1 CA for 100Ah cell is 100A, 0.5CA for 100Ah cell is 50A

edit: updated numbers
 
Last edited:
Yes, but you have to have enough load for cell capacity rating. Preferrably 20-40% of amp-hour rating in amps.

The chart works pretty good when scaled to battery AH rating. That is why I included % CA scale in addition to absolute amps for a 280 AH EVA cell., For a 100 AH cell the 0.2-0.4 CA test current would be 20-40 amps of discharge current.

Only when the electrode thickness cell design is significantly different, like a cell designed for high peak current with thinner electrodes, would the amount of terminal voltage slump be different. Thinner electrode cell design have lower terminal voltage slump for given amount of load. Winston cells are middle range for electrode thickness. The amount of terminal voltage slump is very dependent on cell temperature. Below 10 degs C the terminal voltage slump increases for given amount of current. Hot, 30-40 degs C terminal voltage slump is a little less but not too much better.

A heavily used cell will not be able to support moderate to high discharge current without severe drop in terminal voltage.

Matched cells, at similar state of charge should have close to the same terminal voltage slump for the same moderate level of load current, Measuring cell impedance with a meter like YR1035 is a trend indicator but not a guaranty of matched cells.

If you want to keep tabs on your cell quality, periodically measure each cell voltage with no or minimum load and then use your inverter to load battery to 20-40% AH rating in current. You may be able to get the individual cell voltage readings from your BMS bluetooth monitor. After 2-3 minutes under load check each cell's terminal voltage. The closer they are the better the matching. The less the terminal voltage slump between unloaded and loaded the better the cell condition.
 
Yes, but you have to have enough load for cell capacity rating. Preferrably 20-40% of amp-hour rating in amps.

The chart works pretty good when scaled to battery AH rating. That is why I included % CA scale in addition to absolute amps for a 280 AH EVA cell., For a 100 AH cell the 0.2-0.4 CA test current would be 20-40 amps of discharge current.

Only when the electrode thickness cell design is significantly different, like a cell designed for high peak current with thinner electrodes, would the amount of terminal voltage slump be different. Thinner electrode cell design have lower terminal voltage slump for given amount of load. Winston cells are middle range for electrode thickness. The amount of terminal voltage slump is very dependent on cell temperature. Below 10 degs C the terminal voltage slump increases for given amount of current. Hot, 30-40 degs C terminal voltage slump is a little less but not too much better.

A heavily used cell will not be able to support moderate to high discharge current without severe drop in terminal voltage.

Matched cells, at similar state of charge should have close to the same terminal voltage slump for the same moderate level of load current, Measuring cell impedance with a meter like YR1035 is a trend indicator but not a guaranty of matched cells.

If you want to keep tabs on your cell quality, periodically measure each cell voltage with no or minimum load and then use your inverter to load battery to 20-40% AH rating in current. You may be able to get the individual cell voltage readings from your BMS bluetooth monitor. After 2-3 minutes under load check each cell's terminal voltage. The closer they are the better the matching. The less the terminal voltage slump between unloaded and loaded the better the cell condition.

Ive read that connecting in parallel old and new batteries on lead acid batteries is not good but somewhat ok for lifepo4. Is that true?
 
Ive read that connecting in parallel old and new batteries on lead acid batteries is not good but somewhat ok for lifepo4. Is that true?
In parallel, the greatest risk is if old battery develops high leakage or short, overdischarge damaging good battery. It is more likely a lead-acid will go shorted or have high leakage.

Whichever parallel battery has the lowest impedance and least amount of terminal voltage slump vs current is going to dominate the supply current. You need to be careful how load and charge current is shared between parallel batteries so you do not exceed charge or discharge current limits on a single battery.
 
Last edited:
Is there a point where thermal runaway is an increasing risk?
Not worried about losing capacity.
As a cell ages and terminal voltage slump increases, the losses within old cell under load current increases. Additional loss means additional internal heating. Internal heating watts is pretty much terminal voltage slump times cell current.

If you draw too much load current from an old cell you can cause it to heat up so much it shorts across plastic porous separator causing a thermal run away. Good thing about LFP is the result is pretty much just a burst and vented cell. Gases and spitting guts is fairly hot so can give you a nasty burn but unlikely to start a fire like other lithium cell chemistries.

Bad thing, when primatic cells are packed together, is one cell thermal runaway can cause a cascade down the line of physically packed together cells.
 
Back
Top