diy solar

diy solar

LiFePO4 Wiki Entry

svetz

Works in theory! Practice? That's something else
Joined
Sep 20, 2019
Messages
7,199
Location
Key Largo
Based on some recent discussions put together a wiki page for LiFePO4. As I'm not by any means an expert so would appreciate all suggestions from those of you that are to fix it up, Ideally it's complete, concise, clear, and has references if someone wants to dig deeper. Thanks!

 
A few suggestions

They last for many cycles, unlike lead they don't suffer from Peukert's law
They do have a non 1 Peukert exponent, especially as they near end of life. It starts at nearly 1.01 (nearly zero effect), but can rise to 1.03 as the cells age. This is pretty minor, and only reduces the pack capacity a few percent at higher rates. 1.0 being no Peukert effect.

Overcharging will damage them​


I suggest reformatting this section. Almost all chargers use the term float charging. LFP do no need, and do not benefit from float charging. Float charging should only be employed when a pack is in continuous usage. For example off grid power systems. Batteries which see significant periods of float service typically benefit from voltages around 3.35Vpc. This keeps them around 100%, and reduces stress.

It may also be useful to add a paragraph about absorb voltages and durations. Unlike lead LFP does not need a long absorb duration. Typically a few minutes at the absorb termination voltage is sufficient to ensure the target SOC is reached.


Don't forget to mention the best feature of LFP! They do not suffer from partial SOC degredation. A week, month, year at an SOC less than 100% won't cause them to lose capacity (like lead does due to sulfation).
 
Last edited:
Thanks Luthj! Great suggestions! I'll work those in. There's already a wiki on charge voltages and rates for specific chemistries so no reason to delve into that.
 
I have a violently strong objection to the memory effect portion.
I did too when I read it!

Bur the article you linked says it does have the memory effect:
Ever since lithium-ion batteries started to be successfully marketed in the 1990s, the existence of the memory effect in this type of battery had been ruled out. Incorrectly, as this new study indicates.

...It does not strand capacity....It alters the SoC to voltage response by... 3mV/cell.
What the 2014 paper doesn't show (that the 2010 did) is the voltage bump grows slightly with each repetition. Obviously if that voltage bump grows higher than the max charge voltage (e.g., cycling around 90%), the charger can never erase the memory effect because it stops at the bump voltage. So "fully recharging" the battery leaves you with stranded capacity.

It looks like the 2014 paper revolves more around inaccurate SoC calculations based on the bumps.

What's not really clear is that 3mV/cell. The 2010 paper didn't indicate there was any topping out to the bump voltage, or what the grow was per cycle, and this one didn't mention the effect at all.

I certainly agree with this:
...[misinformation] will only spread FUD.
So, it's important we get this right.

@Steve_S suggested some Google-Foo for more recent papers after 2018 ... possibly someone will run across one that clarifies things.
 
Last edited:
What the 2014 paper doesn't show (that the 2010 did) is the voltage bump grows slightly with each repetition.
I have observed this with a couple small packs which were used in low C applications, sadly none of them were instrumented/logging, so its all hearsay.

There is some rumors that Telsa is may also seeing this with their China made LFP packs. Part of the reason they suggest a 100% charge every so many cycles, and are having issues with range accuracy?

The small, but apparently real bump makes accurate SOC determination in the middle-top of the SOC range difficult, and makes routine charging/discharge to this area (by voltage control) drift down over time.

However this memory effect is easy overcome with higher charge (Greater than 3.45Vpc?). Since most everyone is charging to that level or higher, its rarely seen in the wild.

There are some folks, who in pursuit of mystical cycles, are charging to 13.8V. Assuming they aren't using most of the capacity, they may never notice that their SOC ~13.8V is drifting downwards. So after hundreds of cycles to 13.8V (or less with poorly calibrated equipment), the battery is 10% lower in SOC at charge termination (just a guess). Of course just upping the charge voltage for a single cycle is typically not a big deal (other than possible balance issues after that long).

Anyways, while not commonly encountered, and not permanent, this seems to be a fact of LFP chemistry. Its a bit beyond me, but I guess the lithium ions release energy from the electrodes is not consistent, and depends on the factors present at the time they are moved. Which we already knew, as charge rate, speed, and temperature has some effect on the reciprocal half of the cycle.

It is important to stress that this effect is not analogous to Nickel chemistries where the memory effect is quite deleterious and pronounced.
 
Last edited:
I did too when I read it!

Bur the article you linked says it does have the memory effect:



What the 2014 paper doesn't show (that the 2010 did) is the voltage bump grows slightly with each repetition. Obviously if that voltage bump grows higher than the max charge voltage (e.g., cycling around 90%), the charger can never erase the memory effect because it stops at the bump voltage. So "fully recharging" the battery leaves you with stranded capacity.

It looks like the 2014 paper revolves more around inaccurate SoC calculations based on the bumps.

What's not really clear is that 3mV/cell. The 2010 paper didn't indicate there was any topping out to the bump voltage, or what the grow was per cycle, and this one didn't mention the effect at all.

I certainly agree with this:

So, it's important we get this right.

@Steve_S suggested some Google-Foo for more recent papers after 2018 ... possibly someone will run across one that clarifies things.
It is easy to say that top balancing resets the memory so all of the states of charge are even.
Or that voltage doesn't represent state of charge because there is a memory effect.

If you don't call it memory what do you call it?
 
Last edited:
There is some rumors that Telsa is may also seeing this with their China made LFP packs.
I saw the 2013 paper was concern about EVs, but I don't get that at all. As the engine consumes a lot of power it's not like regenerative breaking is going to cause microcycles over a specific SoC. It shouldn't have anything to do with China, from the 2010 paper it's an artifact of all lithium chemistry except lithium titanate.

Part of the reason they suggest a 100% charge every so many cycles, and are having issues with range accuracy? The small, but apparently real bump makes accurate SOC determination in the middle-top of the SOC range difficult, and makes routing charging/discharge to this area (by voltage control) drift down over time.
Now the accuracy bit makes sense, hadn't thought of that and it makes the 2013 paper's concern make more sense too. Thanks! ;)

There are some folks, who in pursuit of mystical cycles, are charging to 13.8V. Assuming they aren't using most of the capacity, they may never notice that their SOC ~13.8V is drifting downwards. So after hundreds of cycles to 13.8V (or less with poorly calibrated equipment), the battery is 10% lower in SOC at charge termination (just a guess). Of course just upping the charge voltage for a single cycle is typically not a big deal (other than possible balance issues after that long).
I agree. Since the effect is wiped out by a full charge, I think the only "real" danger is to people that set their BMSes to max charge to 90% or higher to make them last longer. I can see over years that bump growing (assuming it doesn't max out) where all of a sudden it's at the max charge voltage and then wham... you've lost 10% of your battery capacity.


It is important to stress that this effect is not analogous to Nickle chemistries where the memory effect is quite deleterious and pronounced.
Agreed!
 
...If you don't call it memory what do you call it?
I think the general concern is the FUD that might be caused from experiences with other chemistries (e.g., NiCad) where the bump was more pronounced and occurred with far fewer cycles. AFAIK, the effect is so small nobody even noticed this until they started asking about how micro-cycles affected the longevity of a cell.

It is easy to say that top balancing resets the memory so all of the states of charge are even.
Just to be clear, top balancing refers to balancing charge in a group of cells. This memory effect is something that occurs within a single cell.
 
I think the general concern is the FUD that might be caused from experiences with other chemistries (e.g., NiCad) where the bump was more pronounced and occurred with far fewer cycles. AFAIK, the effect is so small nobody even noticed this until they started asking about how micro-cycles affected the longevity of a cell.
I was hoping for a new word I could use in place of memory that made sense. The word memory is already used apparently.
Just to be clear, top balancing refers to balancing charge in a group of cells. This memory effect is something that occurs within a single cell.
I meant - top balancing makes all of the cells state of charge even which is basically wiping out the memory effect. Or setting the memory of each cell to start at the same place. That way anything that effects the memory from then on effects all cells evenly so the states of charge stay in sync.
 
Looks like this 2018 paper provides an explanation for the memory effect, they also discovered something new:

While the many-particle model has reproduced many of the unique phenomena in LiFePO4, in this paper we also identified a new phenomenon, that the voltage profile after relaxation is influenced by the relaxation time, with a larger polarization after a longer rest. We refer to this phenomenon as relaxation induced polarization (RIP). In this paper, we demonstrate this effect and explain its cause based on the many-particle model and the active population.

Still trying to figure out how to answer our basic question (how big do bumps grow). Did see a synopsis elsewhere the bumps were more pronounced by higher rates, but I believe that was in the 2013 Sasaki paper too. Need to read it another 20 times or so, but thought I'd link it up for others.
 
@Arthur, you need to find a new word, what you are describing is something very different than both of the types being discussed/debated here (and its already confusing enough with the two semi-conflicting definitions, adding a third will just further confuse things :oops:).

I think what you are referring to is just imbalance between cells / an out-of-balance battery pack.
 
Here's a 2019 paper:

Phase separating Li-ion battery cell cathode materials feature a well-known phenomenon called the memory effect. It manifests itself as an abnormal change in working voltage being dependent on cell cycling history. It was only recently that plausible mechanistic reasoning of the memory effect in Li-ion batteries was proposed. However, the existing literature does still not consistently reveal a phenomenological background for the onset or absence of the memory effect. This paper provides strong experimental and theoretical evidence of the memory effect in phase separating Li-ion battery cathode materials.

Need to read it a few more times, but this was of particular interest:
It was shown that the memory effect is erased upon lithiation of the electrode after the memory writing cycle, which can be achieved by either a very low cut-off voltage or a longer voltage hold at a cut-off voltage below the voltage of the lower spinodal.

Sadly, as far as I can tell so far, the studies aren't looking at hundreds of bumps in succession...they're looking at one or two and trying to figure out what's going on. Walk before you can run sure, but the little voltage bumps in their graphs aren't very exciting. Also unclear to me are if small particle sizes are migrating there, or if the actions are creating the small particle sizes.
 
  • Like
Reactions: Dzl
@Arthur, you need to find a new word, what you are describing is something very different than both of the types being discussed/debated here (and its already confusing enough with the two semi-conflicting definitions, adding a third will just further confuse things :oops:).

I think what you are referring to is just imbalance between cells / an out-of-balance battery pack.
Thanks Dzl. I will try to keep my thoughts to myself. Nobody understands me anyway
 
  • Haha
Reactions: Dzl
Back
Top