Harvested cells analysis.

[Cherry67]

@wolf
If you need extra space, come over to the Netherlands, I have a very nice big shed, you can make youre own corner there.

I am not difficult with rent, 500 cells per month(-:

100kWh
My shipping container will arrive in a month or so.

Wolf

two for me, please.

 

[Wolf]

.....what I can do is take a random sample of 4 harvestedcells of different manufactures and repeat the test. Would you say that is a good start?

I think thats a good start for sure!

OK here arethe results of 4 random cells tested 12 times.
It shows me the samsung needs to be binned
Other than that the max difference between the testers was 8.87% Personally I think that is an acceptable margin of error.
What do you guysthink?
Should I keep on doing these tests?

Wolf

 

[OffGridInTheCity]

On different aspect (please correct me if this is not appropriate) - I have been mulling over the long view and wondering how to track degradation in my cells as the 'years' go by.

Background:
My system is 48v@780ah / 40Kwh. I need 24kwh (60%) PV array storage in thebattery to make sure I don't loose any PV array power to 'battery full'. The batterybank isalittle over 5,000 18650 harvested cellsand its 'live' in that it charges/dischargesmostevery day.

The system automatically turns the inverter(load) on at 52v (3.71v/pack), will top out at amax charge of 56.4v (4.03v/pack - about 88% SOC), and inverter off is at47.5v (3.39v/pack - about 28% SOC). This is a max of 60% DOD in the 88% to 28% range of battery curve - but looks like it will average about 48% with bottom of 28%.


My question:
has been...how long will this battery bank continue to deliver 24kwhon those max PV array summer days. 5yrs? 10yrs? 20yrs?. So the key thing I'm interested in (as the years go by) is 'how much capacity' has been lost due to using the batteries. In theory, once I've lost over 40% of the battery's original capacity - I will no longer be able to get 24kwh out of - even if I charge to 4.2v and inverter off at 3.0v.

Solution:
To measure this, I've settled on tracking AH/volt discharge after the PV array input has gone to 0 for the day - e.g. to ensure I'm tracking 'pure discharge'. I'm using Batrium shunt AH and Batrium shunt battery voltage as the input data. So far (week's worth)theinitial numbers show an average of 67ah/volt. I should see thisdecrease as a loss of battery capability takes place over time - e.g. 67ah/volt should go down to 60ah/volt indicating a 10% drop.

I'm interesting in any comments on how to track battery degradation over timeon a live system (e.g. don't want to shutdown a pack or batteryto do discharge tests etc).

Thanks for letting me share

 

[Generic]

AFAIK, there is no super accurate way to determine battery degradation on a live battery pack. As mentioned in other threads, coulomb counting is a good way to track state of charge, and you can kind of extrapolate degradation based on voltage differences at different states of charge over time. And Wolf will probably tell you that you can probably calculate battery degradation based on the IR of the battery as time goes on (I'd agree with him on that). I think the best way is to periodically fully charge and fully discharge (to 2.8 or 3.0) the whole battery bank and record capacity. Make sure to record ambient temperature, too, as I've noticed that can affect capacity. I can tell you from my limited experience that a periodic full charge and discharge will not have a huge impact on cycle life. As long as you don't stay at 4.2V or 2.8V too long, the impact will be minimal.