Hi Guys, 

I have a question about the cycle count on batteries and the remaining Ah's.

 

I have 2x 2.4kw pylontech batteries (100Ah in total between the two) which have done 200 cycles each, so still quite new. However, I noticed my battery monitoring software tells me that the full capacity of my system is 98Ah of 98%. Its only a 2% loss from full capacity but if my calculations are correct then I will be at 90% capacity after only 1000 cycles. These batteries are rated for around 6000 cycles.

 

So I would just like to know if this rate of regradation is normal?

This article addresses your question, I believe. It is the supposed industry standard of what the claimed cycles should deliver

"Lithium batteries are known for their long cycle life of up to 6,000 cycles. After those 6,000 cycles, there will still be 80% of the capacity left. To put that into perspective, that’s 16 years of daily cycling of the battery."
https://cleversolarpower.com/lifepo4-voltage-chart/

 

29 minutes ago, WernerM said:

So I would just like to know if this rate of degradation is normal?

Your rate of "degradation" is exactly in line with my expectations.  Let me explain:

It is extremely difficult to estimate remaining capacity in a battery.  In practice it can be measured in a laboratory environment by discharging at some set rate and coulomb-counting the output until it reaches a predetermined voltage.  So, what Pylontech has done is a fudge:  They decrease the reported capacity by 0.1% for every 10 full cycles of discharge. (Note that both the 0.1% and the cycle are based on the original capacity)

I have 8 US3000Bs that are extremely pampered - never charged or discharged at more than 0.2C, never below 20C or above 30C, never discharged below 30% SOC, never kept at 100% SOC.   Despite this each battery reduces it's reported remaining capacity by 0.1% exactly as it completes 10 more cycles. 

So, it appears that this 0.1% per 10 cycles is the baseline.

What I do not know is:

• Will the reported remaining capacity be reduced faster if the battery is abused?
• Will the rate of capacity decrease become lower as the battery ages?  If you look at the curve from the US2000 performance data (supplied by Pylontech) the curve flattens out a lot with age.

I find it an interesting topic.
I have 2 Pylontech with about 400 cycles and 2 with about 300. The reported SOH is 98% and I have the same doubts.
But I think I've read about people who upgraded the firmware and the SOH health status increased spontaneously.
This makes me wonder if my batteries are still working fine with the firmware that came from the factory, should I update their firmware...? or those that do and the reported SOH increases, is the value obtained since then unreliable?

The SoH is interpolated based on what they expect. Unless there is a big jump, don't read too much into it.

15 hours ago, Cef said:

But I think I've read about people who upgraded the firmware and the SOH health status increased spontaneously.

If my memory serves me correctly the formula that I describe above came in when I upgraded my firmware from 1.7 to 2.8.  That actually decreased the reported SOH.

The new formula would give you 97% SOH at 300 cycles and 96% SOH at 400 cycles.

Hi @Calvin

I admire your conclusions above all because of the way you say you take care of your batteries and appreciate you clarifying this issue.
My case is:

US3000C
2 purchased in the first days of December 2021 (approx. 400 cycles)
2 added in the last days of March 2022 (approx 300 cycles)
My system reports SOH = 98%, it may be about to change... ?

On 1/27/2023 at 9:58 AM, Calvin said:

So, it appears that this 0.1% per 10 cycles is the baseline.

What I do not know is:

• Will the reported remaining capacity be reduced faster if the battery is abused?
• Will the rate of capacity decrease become lower as the battery ages?  If you look at the curve from the US2000 performance data (supplied by Pylontech) the curve flattens out a lot with age.

 

Will the reported remaining capacity be reduced faster if the battery is abused?

For everything published about batteries it should be like this... The more violent the discharges and charges, the faster they should degrade. Temperature is also a fundamental factor that rarely one takes into account.

Will the rate of capacity decrease become lower as the battery ages?  If you look at the curve from the US2000 performance data (supplied by Pylontech) the curve flattens out a lot with age.

I really hope so. If your formula is correct, I don't know how you could reach approximately 80% SOH with 6000 cycles... since if it were a linear function at 6000 cycles you would have an SOH of 60%

 

 

 

10 hours ago, Cef said:

Temperature is also a fundamental factor that rarely one takes into account.

Absolutely.  The graph below is from the same Pylontech document.  My batteries live in a very well insulated compartment with their own air conditioner. The aircon of course runs only on hot sunny days when I have plenty of spare power.

 

10 hours ago, Cef said:

If your formula is correct

Remember this "formula" is simply an observation of what happens on my US3000B batteries on firmware 2.8 - it may be different for other models or firmware versions.

As for how I observe this: I have custom management software that interrogates the BMS (every 2 seconds) and reports the cycle count (which is available to about 5 decimal places) every time the SOH decreases (which it does in 0.1% steps).  The correlation is perfect - each 0.1% decrease happens exactly at 10 extra cycles.

 

10 hours ago, Cef said:

since if it were a linear function at 6000 cycles you would have an SOH of 60%

Actually 40% (100-6000/100), which is why I expect that their curve must start flattening out at some point.

13 hours ago, Calvin said:

Actually 40% (100-6000/100), which is why I expect that their curve must start flattening out at some point.

Yes @Calvin that's right!, sorry, SOH reduction of 60% would leave it at 40% if it were a linear function.

On the other hand, is it your own software development? or do you use ICC ?

14 hours ago, Cef said:

On the other hand, is it your own software development?

Yes.  With a lot of help from information provided by people like @Coulomb and @Youda

1 hour ago, Calvin said:

Yes.  With a lot of help from information provided by people like @Coulomb and @Youda

I'm currently using the MultiSIBControl software for my solar monitoring, works well and helps me tremendously. 

 

Are you using something similar?

Thanks @Calvin !
I'm going to investigate the communication protocol to incorporate it into my own development. Added to task list...

4 hours ago, WernerM said:

Are you using something similar?

Yes and no.  Functionally my system is quite similar. The architecture though is different - my system is split into 2 layers.  The control and data acquisition part is written in C++ and runs on a 32bit ARM microcontroller.  The human interface (settings, graphs, history) is written in C# and is integrated into my wider home automation/security system.