LiFePO4 Best Practices

My understanding is that the problem to only take it up to 80% is that a cell imbalance would start to happen. Unless you have an active balancing BMS. That is why most people fully charge, so that balancing can occur, and then just not use all the way down past 20%.

Temperature is a big culprit, so keeping those discharge and charge amps low and not keeping the battery at 100% SoC while it is also very hot would be a good start.

I’m actually considering getting a third battery, not so much for the capacity, but that it’ll bring the amps drawn from each battery down.

Individual cell balancing issues only manifest themselves at both ends of the SOC curve. No need to balance the cells if they don’t get to near full or empty. The cells have to be somewhat balanced, by this I mean that it would not be a good idea to put a fully charged cell in series with a fully discharged cell and expect that by magic the cells will work to deliver 80% of the capacity.

I think if you charge up your cells so they are all at 3.6v by whatever means, all parallel to 3.6v +0/-(50-100mV) then let them sit there to balance themselves or in a string with a BMS to balance them. Then set your control system to only allow the inverter to draw between 20-80% SOC you will not have a balance issue for months of charge discharge cycles.

For the hobbyist it would be interesting to watch the cells slowly go out of balance but most people don’t care. Just use a BMS that will not take an individual cell into a damaging condition and you will not have to think about it.

It is only for people trying to extract the maximum capacity out of their cells that top balancing on every cycle becomes important. If you don’t try to charge them beyond 80% SOC you probably will not see a difference for a fair while.

Interesting. Is it actually possible to ask the BMS to only allow charging up to 80%? I know you can obviously limit the SoC at the bottom, but haven’t seen the option to limit SoC at top.

Would taking the battery to 100% still be an issue if you manage to keep it cool while doing it?

Heat is definitely a sign that you are stressing the cell. You can limit the amount of energy that is converted to heat by not charging too fast and by avoiding the extremes of the charge curve. What do you want to achieve beyond extending the life of the battery. I want my cells to be sufficiently charged so they will hold up overnight until the sun rises again. I therefore need a certain amount of charge to be put into the cells. I know that I will get more life out of the battery if I charge it at 0.2C but I intend to limit my charger to 0.5C and see how that goes. I do want to charge my battery in the day light hours and that will require more than a 0.2C rate unless I have such a large battery that I can do that and still have enough to provide the energy I need.

To recap:
Recently gotten hold of 2nd Life cells, having had an immense difficulty as GRO said, getting them to and keeping them balanced at a SOC of >95%, I did the bottom balance.

Reed kept on telling me to bottom and top balance, vid after vid he sent.

The Bottom Balance method I followed:
Drained all the cells to 2.8v, inverter got them to like ±3.1v, rest I took out using “The Bulb V2”.
Then I connected the cells in parallel for 48 hours.
After that I recharged them to 100% SOC, watching how they stayed in line all the way up to >95%.
At >95% I then helped a cell or two with “The Bulb V2”, and once things calmed a bit …
I discharged the bank down to 30% again with normal use.

Since then I seem to have a between 0.004 - 0.015v difference at 100% SOC, with the cells keeping in very close proximity as they recharge from SOC of 30%

It was fun, tedious, but fun.

Today I can say with the utmost conviction:
If you want to know what your cells are doing all the time, get a BMS that can show one the individual cell voltages all the time via i.e. one’s phone, whilst the BMS is interfaced with the system/Venus.

True. You don’t have to balance them every single day. Usually you only need to balance after prolonged periods of not being fully charged, which could be months. If you had a deep discharge event, then you may observe that balancing is required again.

And it also depends a lot on the kind of balancer. If it is an active balancer (A DC/DC converter that steals from the rich and gives it to the poor) can technically balance the battery at any point, it can even bottom balance it. But active balancers are more expensive, so most batteries use a passive balancer that only bypasses the high cells, which means 1) it only balances at the top, 2) it only balances once one of the cells exceeds some threshold, and 3) there must be actual current flow into the battery for the balancer to have something to work with.

This threshold is usually somewhere above 3.45V per cell, I believe for Pylontech it is 3.485V per cell (empirically determined by a user on the “other” site :slight_smile: ). At this point the battery is usually well above 85% SoC.

My most recent experience with BYD suggests that they “count electrons” using a shunt or a hall sensor up to 99%, but the SOC will only go to 100% once at least one cell spikes up in voltage. Since SOC drift is also a real factor for batteries that use a single hall sensor (BYD, BlueNova), prolonged periods at partial charge may cause other issues too.

I’d say, to be on the safe side, charge it up at least twice a month.

Also consider warranty conditions. Some manufacturers will deny a warranty if you didn’t regularly charge the batteries. Eg Victron SmartLithium batteries must be deployed with “BatteryLife” turned on. This is a feature that ensures they get fully charged at least once every ten days… :slight_smile:

1 Like

Fully support checking on your batteries periodically. Once you get into a rhythm you will know if you can extend the interval between checks. You will also get a feel for what sort of events will result in some action being required.

1 Like

Imho if you stay in a place like Sutherland… and it gets -12 in the winter you might cause permanent capacity loss when charging your batteries below 0. I am not sure if sub zero is such a big issue in most of South Africa as most people will keep the batteries indoors and the temp should be above freezing point. If you keep your batteries in the garage and there is enough cold air that can pass through the garage door to drop the ambient temperature this might be something to monitor. Pylontech seems to have low-temp cutoff so at least they will protect themselves. Not sure about the other brands. My DIY battery Bms does not have any temp sensor so I will have to keep this in mind during winter

2 Likes

I think Victron SmartLithium batteries stop charging below 5°C. It depends on the cell as well, I understand that some chemistries are better. For example LiFeYPO4 (with Yttrium) cab handle heat better, but are worse in the cold.

Most commercial batteries will raise a charge-temperature warning, and if you don’t heed that, it will protect itself by blocking charge or disconnecting from the DC-bus.

Sorry to resurrect this one.

I was looking through VRM and I noticed the battery temperature widget:

I have 3 Pylontech US3000’s; they’re about a year old. Should I be concerned that the temperature is basically hitting at least 30°C every day and going over and beyond it during the daylight hours?

It don’t put a lot of load on the batteries during the day and didn’t feel like the location where they’re mounted (garage) is that hot either…

I don’t think so, no. I think 30°C is completely normal and even on the low side compared to others, but I may be wrong. Looking at a friend’s new install his US3000’s are hitting 32°C, so I would say it’s normal. The temperatures will rise when charging and discharging.

The datasheet also mentions a working temperature of -5°C to 60°C. The Pylontech BMS also monitors temperatures and will shut down if there is a high temperature event to protect the cells.

Yup. These go up all the way to 60°C when things go really wrong. The BMS will reduce the charge/discharge currents if you go over 40°C. Before then, there is nothing to worry about.

For interest sake, you should plot the delta between that and ambient (will probably take a bit of work). Remember that ambient itself can swing 10°C a day easily, and with ambient at say 24°C (pretty typical for my area), 30°C means a mere 6 degrees…

1 Like

Didn’t lab tests show that if LiFePO4 batteries are above 25 degrees C and in a high state of charge, it hurts the capacity of the cells? As in, a high state of charge for a year at 30 degrees vs a high state of charge for a year at 25 degrees has significantly different impacts on the deterioration of the cells.

Will see what I can do. I need add a temperature monitor to my garage first for ambient temperature.

Wow, guys I’m new here, but very thrilled to see how much info Is here.

Reading to catch up…Too little time :smiley:

87