Hi,
Does anyone have any studies, reports on how the longevity of LiFePo4 batteries is affected by different ranges being used?
e.g. for longevity, is
0-100%
20-100%
0-80%
20-80%
best, and by how much?
A lot of people seem to have opinions on this but was wondering does anyone know of any proper research here?
Many thanks,
Mark
There are a lot of research papers about for LFP batteries but they tend to focus on specific things like ageing, or temperature - the main problem being that LFP batteries do not have enough historical data to study as the chemistry is fairly new.
This is a very good video that explains some of the realities
The main thing to remember is that Fox have already limited the batteries to 90% DOD and they provide warranties for their batteries being charged by their BMS systems which do require regular 100% charge cycles to maintain SoC accuracy. Best advice is to use them within the Fox specifications temperature, currents and their battery warranty will provide all the guarantees you need.
This is a very good video that explains some of the realities
The main thing to remember is that Fox have already limited the batteries to 90% DOD and they provide warranties for their batteries being charged by their BMS systems which do require regular 100% charge cycles to maintain SoC accuracy. Best advice is to use them within the Fox specifications temperature, currents and their battery warranty will provide all the guarantees you need.
Thanks Dave!
That makes sense!
when you say:
"which do require regular 100% charge cycles to maintain SoC accuracy"
Any ideas how often 'regular' is?
1x per month
1x per week
1x per day?
That makes sense!
when you say:
"which do require regular 100% charge cycles to maintain SoC accuracy"
Any ideas how often 'regular' is?
1x per month
1x per week
1x per day?
It’s best to do it monthly, it won’t damage them if you don’t but you’ll see the accuracy of the SoC gets worse with occasional jumps as it re-evaluates the Soc from the cell voltages.
I found this one quite interesting
https://iopscience.iop.org/article/10.1 ... 111/ad6cbd
There's also https://www.sciencedirect.com/science/a ... 1722000283 and https://www.mdpi.com/1996-1073/14/6/1732 which cover aspects like calendar degradation.
Personally I do 100% once a week, 10% once a month via an automation. It'll do it more often in certain circumstances (for example, if forecast PV for the day is very low, it'll use the full range instead for a bit of extra arbitrage to make up for the lost generation).
It's not worth doing the calibration cycles manually, you might as well just stick to the full range rather than trying to keep track. Also, underutilising the battery is just as much of a waste, due to calendar degradation (for example, sticking to 20-80 all the time, 0.6 full cycles daily, would mean you'd run out of warranty years, and the battery capacity would have decreased anyway due to the passing of time). It's a bit of a balancing act.
Calendar degradation is (relative to cycling) more important on LFP, whereas with NMC batteries the cycling is a much bigger factor.
https://iopscience.iop.org/article/10.1 ... 111/ad6cbd
There's also https://www.sciencedirect.com/science/a ... 1722000283 and https://www.mdpi.com/1996-1073/14/6/1732 which cover aspects like calendar degradation.
Personally I do 100% once a week, 10% once a month via an automation. It'll do it more often in certain circumstances (for example, if forecast PV for the day is very low, it'll use the full range instead for a bit of extra arbitrage to make up for the lost generation).
It's not worth doing the calibration cycles manually, you might as well just stick to the full range rather than trying to keep track. Also, underutilising the battery is just as much of a waste, due to calendar degradation (for example, sticking to 20-80 all the time, 0.6 full cycles daily, would mean you'd run out of warranty years, and the battery capacity would have decreased anyway due to the passing of time). It's a bit of a balancing act.
Calendar degradation is (relative to cycling) more important on LFP, whereas with NMC batteries the cycling is a much bigger factor.
Thanks WyndStryke for your thoughtful reply (and for the papers), much appreciated! 
What SoC range do you typically use for your batteries then?
What SoC range do you typically use for your batteries then?
It changes on a day to day basis, anywhere from 14-92 to 20-84 (my automations will tweak it based on the PV forecast / weather forecast / etc), but I will also be cycling in the middle of that range as well (again, dependant on things like cell temperature / PV / weather / etc). I also adjust charge and discharge rates based on those criteria. On very hot days etc it will be using the slowest possible charge and discharge rate and no mid-range cycling (a high cell temperature causes increased cell degradation). When it is cold, it will be more aggressive, to generate warmth.What SoC range do you typically use for your batteries then?
So I'm using more like 0.8-1.0 full cycles daily even if I don't hit the extremes. That's only possible on particular tariffs.
I would honestly expect FoxESS to manage this properly. The only thing as a enduser I should set is the MinimumSoC and the rest the system should "automagically" manage.
System Setup (Austria): Inverter: STAR-H3-12.0-E / Battery: STAR EP11 (10.36 kWh) / Panels: 24x Canadian HiHero 440 Wp (10.56 kWp total)
That's how 95% of people do it. I tend to go above&beyond.