<quote>SOK sk48v100. i don't think it's smart bms.</quote>
It has the smart bms actually. I didn't found yet the info which exactly BMS that is, but the battery has communication ports so most probably there are some apps (for phone or for a PC) are available to connect to it for monitoring and hopefully at least see what settings are pre-set there (exact voltage of the "overvoltage cut-off" per cell is most interesting) and may be even able to change them (if it is above 3.65/per cell I would definitely bring it down to 3.65). Also "balance start voltage" (it considered to be better set it to start at 3.4-3.45 per cell), although with under 1A balancing current it can't help much if your charging current is higher) ; The most useful info about that battery I just found out is that "Cells are not welded (user replaceable)"
No, don't worry, I'm not going to recommend you to change them but that's a good sign because if worse come to worse you can easily (and very usefully) re-arrange them to parallel connection to do real practical helping-and-fully-solving-the-problem-to-fully-top-balance-them using a 3.65V CC/CV charger, not a big problem if that charger will be even 5A max current so it could take long time for 16x100Ah cells (but we will do estimate calculation how long it can take potentially) but it will be significantly better anyway rather than expecting the BMS to balance them within "10-15 cycles"....
I read this text from the user manual which you quoted and it seems that those user manuals are not the best source to learn about LFP batteries. Let's see what they are saying: "From the factory, the charge level within each cell differs slightly", "initial balancing process is slow and typically takes about 10 to 15 charge cycles" - ok but wait a moment: you are a manufacturer, you building complete ready battery, why you didn't perfectly match and top balance your cells before put them into that battery? ; "(a difference) ... can reach as high as 250mv. This is normal and no cause for concern" - NO, it's definitely NOT NORMAL ! too much to me ! (Ok, may be it's still useable with that difference but why leave it like that if we can easily bring that difference to as low as 0.005V ? What means the "250mV difference" is that while one cell is already 3.650V the other one can be still at 3.400 V ? too far from "normal" I guess...
Then what other thing they wrote was "OVP triggers and BMS will disconnect charging and then "balancing will start" ? What? Actually the balancer should "start" yet before that - under charging yet after cells reached 3.4 - 3.45 each... But if you want that balancer (under 1A its max balancing current) be able to balance then total charge current should be not much higher than the same 1A as I understand; otherwise see: it trying to "pull" let's say 1A from a cell but total current goes INTO that cell is 10A so how much that can help balancing them? anyway when they are near fully charged the voltage on a cell rises too rapidly - no time to top balance evenly. I was trying to balance cells that way in my self-assembled 4S pack - never achieved even with an active balancer 5A max balancing current!
Anyway, we can still perfectly use the battery with charging it up to 95% (setting "absorb" (i.e. max charge voltage) to 16x3.45V) and with "nothing to worry about" if cells are really have the same capacity but just only small difference in their SoC (=small voltage differences between cells at their almost full SoC, an example of not much is 30mV, i.e. : 3.64 -- 3.61 ) (another example is: my cells now at 5mV diff: 3.3601-3.3604-3.3605-3.3606 and that is at 93% SoC according to Ah counter and still under charge current of 7A for 280Ah capacity but I know surely they are still not perfectly top balanced ! If I bring them to 3.65/cell then some would reach 3.65 while some others will be still at 3.58). Of course if cells have different CAPACITY then we can charge the whole pack to the max capacity which is not more than the CAPACITY of the lowest cell (not by voltage or SoC - but by real its full Ah CAPACITY) - but that should not happen with cells inside of a manufacturer-built battery! to me , even the SoC difference should not happen more than by 30mV difference between cells at near fully charged state.
And you are absolutely right , why to "waste" (for nothing in return) expensive battery's resource if we easily can use it that way to extend it's life significantly. That's the reason why I'm saying why charge it with high current to full within 1 hour if we can perfectly do the same within 5 hours with lower current. Also: what is most affecting the longevity of LFP? - 1) higher temperatures 2) higher current of charge/discharge 3) longer time under higher charging voltages. All of those can be EASILY avoided with no practical disadvantage.
1) 25-27 degrees Celsius is OK and I guess up to 32 ambient temperature is still OK but higher is not so good as I remember,
2) it's perfectly normal the charge/discharge current up to 0.2C (for 100Ah battery it's = 20A so with "48V" battery it's about 1000 watts) , it's "capable" of enduring up to 1C current but that is what significantly reducing the cycle-life
3) much better allow charging up to 3.5 per cell rather than always to 3.6 per cell (the difference of SoC between 3.5V and 3.65V is very little because after 3.55V under charging the voltage rises very quick!). I would use above 3.6V per cell ONLY for top balancing.)
So my recommendation to you is: yes, try to do , as you said, "cycle them several times and see what happens. i do need to try and keep better track of the cell voltages so i can see if they are in fact getting closer." and see , but I guess it will not anyway perfectly top balance if was significant mismatch of SoC between cells from the beginning (if the manufacturer didn't done their "homework" good enough
then no problem again but in that case the best thing to do is to connect cells in parallel and do real perfect top-balance using 3.65V CC/CV charger (power supply). After that be perfectly fine for the next 5 years and YES, definitely under normal usage that batteries should last minimum 10 years until their real capacities start declining to about 80% of their "new"-state capacities.
Estimate calculation of time needed to top-balance using 5A max current CC/CV charger for 100Ah x 16 cells: When calculating that we can take into account that before connecting them in parallel to that charger they will be already pre-charged up to about 90% even more. So it's only 10% to charge that way. So 10Ah left. 16 x 10Ah = 160Ah with 16cells connected in parallel. 160/5A = it will take max 32 hours estimated. Not so big problem anyway. But may be your charger will be even 10A.