LiFePower4 Battery State of Health degradation...

[SeaGal]

Yes there is an actual equation this is not just made up. It's ( actual remaining capacity AH/ Rated AH capacity) for example 98ah/100ah= 0.98%.

Hopefully 98% in that example, though

 

[Johno1066]

The only way that I can think of to determine the batteries SOH is to fully chrge the battery and then fully discharge the battery into a load where you can accurately measure the amphours delivered and then measure that against the original battery capacity. Think your BMS can do that.

 

[JLBinTN]

Assuming it actually means anything, you'll be at 80% SoH @ 4180 cycles. Why aren't you happy?

Literature says 80% SOH at 7000 cycles.

 

[sunshine_eggo]

Literature says 80% SOH at 7000 cycles.

And you believed that?

EV grade cells have a lower cycle life, but somehow these server rack batteries are using non-EV grade and getting longer cycle life?

 

[JLBinTN]

I Wouldn't worry about the SOH reading, until the battery starts smelling like peanut butter.
What makes a battery smell like peanut butter?
Absolutely nothing, unless you smear peanut butter on it.

Good point. I ignore check engines lights too.

 

[timselectric]

Good point. I ignore check engines lights too.

Check engine lights are real.
I wouldn't do that.

 

[cinergi]

SoH definition and implementation is up to the manufacturer. I have an Orion BMS. My SoH was dipping down quite a lot (by 10%+) ... as a result of high IR developing between the bus bars and cell terminals. Once I fixed that, my SoH has remained at 99% (450 cycles so far). From their site, this is what they define as SoH:

Typically this value is calculated based on observed capacity and nominal capacity (that is, a relationship between how much capacity is actually observed by the BMS and how much capacity the battery pack started off with when new) and the Internal Resistance of the pack.

I personally wouldn't pay attention to any SoH differences less than 1%. To really know, do a capacity test.

 

[timselectric]

Yup
A capacity test is the only way to know how your battery is doing.
Unless you fully cycle the battery every day. The BMS has no idea what the actual capacity currently is.

 

[sunshine_eggo]

Good point. I ignore check engines lights too.

A check engine light is activated based on very real and traceable condition well defined by the manufacturer and documented.

You are comparing visible apples to invisible oranges.

 

[Hotwires]

Not sure what to think with my issue. I thought 1 pack being 30% difference compared to the rest would indicate a bad battery. It started trending that way again, then over the last week, now I do NOT see an issue. SOC seems to be holding with the rest of the batteries. Maybe I'm looking at it too close, but hit the panic button when I saw the difference. Pack #2 was the suspect pack, now it seems to be fine. IDK...

 

[fafrd]

SoH definition and implementation is up to the manufacturer. I have an Orion BMS. My SoH was dipping down quite a lot (by 10%+) ...

as a result of high IR developing between the bus bars and cell terminals. Once I fixed that,

Could you comment on how you determined you had high resistance developing between bus bars and cell terminals?

And how did you go about ‘fixing’ that?

my SoH has remained at 99% (450 cycles so far). From their site, this is what they define as SoH:

I personally wouldn't pay attention to any SoH differences less than 1%. To really know, do a capacity test.

I’m starting to plan for a new system which will push my 560W 8S2P LiFePO4 battery harder than it’s been pushed over it’s first 3 years.

Just looking at how much battery voltage drops when charge current or discharge current changes abruptly, I’m suspecting I may have some higher-than-ideal resistance between bus bars and aluminum cell terminals.

So I’m in the market for any advice regarding how to determine which connections may need improving and best practices to do so…

 

[fafrd]

Possibly toward the end, below 50% for example.
Even though it is a different chemistry, my experience with EV batteries is that there is a steep initial degradation of 5-10% in the first year, then they plateau off for the next seven or eight. That is the longest I have owned an EV. I am still using some ten year old Nissan Leaf modules that still have 60-70% of capacity but I have no knowledge of their prior ownership,

So a 280Ah cell might drop to 252-266Ah after year one but should then remain close to that for the next few years?

I’m 3 years in and planning to tune up my battery as appropriate this winter, so interested in any advice for how to do so…

 

[cinergi]

So a 280Ah cell might drop to 252-266Ah after year one but should then remain close to that for the next few years?

I’m 3 years in and planning to tune up my battery as appropriate this winter, so interested in any advice for how to do so…

Different chemistries with different curves and usage characteristics altogether. I wouldn't compare.

1 year into my system (280Ah LFP cells), I tested to 99% of my original capacity.

 

[cinergi]

Could you comment on how you determined you had high resistance developing between bus bars and cell terminals?

And how did you go about ‘fixing’ that?

The BMS made it obvious in several ways ... 1) it shows the calculated IRs of each cell and they were climbing at different rates, 2) it reports State of Health and it was degrading (due to IR), and 3) the voltage graphs of each cell were varying more and more wildly during charge and discharge. Once I re-did all the cell connections, the graphs/values returned to normal. In my case, the cause was having solid busbars connecting cells that are in a spring-compression fixture.

I’m starting to plan for a new system which will push my 560W 8S2P LiFePO4 battery harder than it’s been pushed over it’s first 3 years.

Just looking at how much battery voltage drops when charge current or discharge current changes abruptly, I’m suspecting I may have some higher-than-ideal resistance between bus bars and aluminum cell terminals.

So I’m in the market for any advice regarding how to determine which connections may need improving and best practices to do so…

Without a baseline it's hard to know. Many variables (cells, busbars, terminal type, clamping force, fuse, wiring, BMS, etc). If you're unsure, re-do the connections -- clean the busbars and terminals (plenty of threads here on that) and immediately reconnect and test. Make that your baseline if it's any different than the numbers you're getting now.
It's not unexpected to see a decent voltage difference between charge and discharge even inside the flat area of the charge curve especially depending on where you're measuring.