LFP longevity factors

[john61ct]

A thread for people to discuss their beliefs about how to make LFP last much longer than the manufacturer cycle ratings.

Link to / from other relevant threads.

Other than standard rules like Be Nice,

Feel free to say what you like.

Even, "it's just not possible!"

Or not worth worrying about.

Or since there's no published research, we should not experiment as users or even professionals.

Speak your mind!

[kmacdonald]

They will be destroyed by an error or failure long before they wear out. I never heard of anyone wearing a LFP bank out.

[rgleason]

I've noticed this factor too. Something happens to drain them down to 0% or they get overcharged, shortening or ending the lifespan. This is why the system needs to be designed to be fairly automatic, or with the proper alarms (at least for me, because when sailing, my focus needs to be on navigation and sailing.

Originally Posted by kmacdonald

They will be destroyed by an error or failure long before they wear out. I never heard of anyone wearing a LFP bank out.

[rgleason]

Endurance LiFePO4 battery testing
S. Papezova1,* and V. Papez2

GWL -LiFePo4 Cycle Life - What is the real cycle life of LiFePo4? Definition.

GWL EV-Power -LFP cells proper charging characteristics - Typical commercial recommendation for charging showing "..constant current charging and rising to the full Voltage, and maintaining full voltage while current is reduced. When the current drops low the charging is terminated"

Many LiFePo users in this thread have more conservative approaches which stop charging at the point where the "full voltage" is reached.

[rgleason]

Winston LFP-100AHA-Wide - Charge/Discharge Charts
------------------------------------------------------

1. Voltage vs SOC for Charge & Discharge show that 100% Capacity (C) is measured at 3C Charge and 3C Discharge.
2. Since most of us do not expect to charge and discharge at that rate, there is more effective capacity available.
3. Voltage vs SOC for Charge and Discharge at .5C show Capacity (C) is about 115% when discharge reaches 2.8v and charge reaches 4.0v. (Most of us would not use those values.)
4. If the high cutoff is about 3.55v and the low cutoff is set at 12.8v then from these Winston tables it appears that 95% of the capacity is available.
5. If the charge rate is held to .3C and discharge held to less than .3C then there would be an additional benefit in capacity & probably cycles.
6. Note that temperature really makes a difference in voltage and capacity. Don't put these batteries in a compartment next to the engine without good fresh air circulation.

[rgleason]

Charging LFP (LiFePO4) batteries to 90% SOC, or not


1. Charge Rate affects SOC. Higher charge rates result in lower end SOC.

2. Charge Rates affect Longevity. Higher charge rates reduce longevity.

[rgleason]

CALB vs Winston LiFePO4 batteries

There are differences between manufacturers, so naturally there will be differences in longevity, however provided they are good known manufacturers, some of the Manufacturer rating may seem to be very different, but they may have different measurement standards.

I may be wrong about this, but I would guess that for the most part, the differences are probably not as great as they many seem to be.

[john61ct]

Sorry been holding back on contributing here, been distracted on other related topics.

Regarding SoC vs longevity, and why I think "20-80%" is a grossly exaggerated expression for the AH capacities lost by "avoiding the shoulders" for the sake of longevity.

http://www.cruisersforum.com/forums/....php?p=2808578

[rgleason]

John, I thought the three types of capacity measurement were decided to be:

1. Manufacturer's capacity.
2. Absorb / CV set at 3.65Vpc, taper to zero current, no amps flowing at all, call that 100% SoC.
3. Absorb is set to 3.45Vpc, and charge stops when trailing amps have tapered to .06C

I imagine you may think that SoC is significantly lower?
It is in reality 98.69%

I think EVM1024 in this Zero Sum vs Sliding Window thread is explaining the concept more than trying to establish some Actual Capacity standard of 20%SOC to 80%SOC.


In the Alternator with Lifepo thread I wrote: (Therefore I am going to assume the max they'll be charged to is 95%, leaving 75% usable.)



Do you think 20%DOD is too high? What is reasonable, considering longevity?

[rgleason]

More about charging and longevity here


Maine Sail wrote

I noted an interesting phenomenon with these over-discharged and over-charged cells. When starting from 2.5V charging at 40A the cell would attain 3.55V within approx 1:45 and then enter a long current taper, just like lead acid does, once voltage was held steady. So yes the cells are ruined despite still being able to deliver approx 91% of their capacity. I did find the recovery of usable capacity interesting and I would love to know what caused that..?

OceanSeaSpray

I think that the capacity recovery and slow current taper you are seeing is a separate phenomenon and some of it could stem from the fact that these cells hadn't been properly charged for quite a while. Some of the lithium may have become more difficult to reach or shift.
The other thing is that they might now have much higher internal resistance than when they were new and that would extend the current taper (I would fully expect that).


Other than that, there is charging and charging. What people don't seem to understand with lithium is that overcharging is not so much a function of voltage, but rather how long you keep at it.


With the old lead-acids, charging can be continued indefinitely as long as the voltage is not too high and it is called charging. LFPs get destroyed by the same regardless of voltage.

OceanSeaSpray wrote:

Voltage determines how much stress is placed on the cells, how early they reach absorption and eventually how long it takes to recharge them, but a simple look at the charge 0CV curve shows that a fully charged cell sits near 13.45V, so anything a little higher than that is going to try and charge the cells even if they are "full".

Conversely, if they are not full, they can't be overcharged and, as long as the charge is terminated well before the cells are full, voltage doesn't matter much.

OceanSeaSpray continues, about Alternator Charging:

Cells charged to 3.6V by an alternator are far from full if charging is terminated as soon as the current starts heading down. In fact, it is difficult to get more than 75-80% SOC this way if the C-rate is high, say 0.5-1C.
The voltage problems arise when the charging regime keeps going. Elevated voltages are very bad, but lower voltages are that good either. Charge termination matters much more than voltage.

So my estimates for %SOC at a charge rate of .35C are not that far off.

Interesting that he mentions
Charge rate of .5C-1C with termination voltage of 3.6vpc (14.4v) reaching 75%-80%SOC,

In "Alternator Size with LiFePo4" post #434 I considered
Charge rate of .35C up to a termination voltage of 3.47vpc (13.88v) reaching 75%SOC. Perhaps I will need to increase the termination voltage, to get to 75%SOC but this is what Maine Sail and others use, it is a slower charge rate and is more likely to reach a higher state of charge than a charge rate of .5C.

Of course when needed, it would be nice to charge at .5C occasionally. This would require some kind of easy Regulator control to adjust the C-rate similar to a Balmar Belt Manager setting, but more easily adjustable. A regulator that allows easy setting of this through a simple control or Android App would be very useful. Perhaps the WS500 will have this ability. and al thomason regulator

[john61ct]

Originally Posted by rgleason

John, I thought the three types of capacity measurement were decided to be

I was illustrating the fact that SoC definitions are arbitrary, and giving labels to show how **different definitions** can be useful in different ways.

How each vendor and owner choose to actually define those labelled setpoints for themselves is up to each.

_______
I was saying that the cliché "20%SOC to 80%SOC" is harmful, because it implies losing 40% of your capacity is required to get decades of longevity. I feel very strongly that even being as conservative as me, is not sacrificing **anything** significant, and therefore it is just silly to keep pushing to those higher levels.

Using the shorthand "avoiding the shoulders" is better IMO, more accurate because it is less precise, and

reflects the huge variability of those curve slopes, depending on how the 0% and 100% are defined, and current rates.

And the **real** fundamental point is, just using SoC% is IMO meaningless - % of what? measured how? needs to be there, not just a % number.

_______
When talking about the bottom end, gaining longevity by reducing average DoD, that is 100% an owner judgment call, no one else's opinion can be relevant.

I say, just make **sure** you never get near 3Vpc in normal usage, and maybe size higher than you need for Reserve purposes.

A mostly solar rig will need **much** more of that than one where the owner can crank up an ICE charge on demand.