LiFePO4 battery capacity decreasing

[Epicurean]

I have a problem with the state of charge of mi Super B LifePO4 house batteries. They seem to have lost almost 0.5% of capacity/day over the last two months, although not at a constant rate. The max I can get now is 67% capacity and then they go into a float stage about 2:00 PM, even though the sun is shinning strongly. I charge them through an MPPT solar charger that does a bulk charge then an absorption stage before going to a float stage until they start discharging at night.

Any suggestions?

Thanks

[senormechanico]

What are your voltage settings and can you be CERTAIN THEY ARE ACCURATE?
A few tenths can make a huge difference.
There are several threads on CF about LiFePo4's.
I have them both in the home backup and the boat and would never go back to lead.

[john61ct]

But they are not a commodity of uniform quality, online market's full of lying cheating scammers.

Please post a link to your purchase source. Who designed your overall system? Who did the installation labor?

And also for the ancillary equipment, charging & protective gear, details on the wiring / terminations etc, diagnostic tools.

For example, how are you conducting your capacity load testing?

Finally, please describe your usage patterns & care regime, start / stop setpoints, Ah per day, energy input sources, how long away from mains at a time.

And anything else you think might help us help you. If what you say is true, and assuming a good quality bank to start with, that's pretty drastic bank murder going on there.

[senormechanico]

How long Absorb at what voltage? What float voltage?
John's asking good questions that need answers.

[Epicurean]

First, thanks for your input!

Today it did an absorption charge at 13.95 volts. It was set to do this for 300 minutes but only lasted for 45 before switching to the float stage – 13.45V. It says it reverts to the float stage if the absorption time period exceeds the user defined time allowed or if the charging current at that constant voltage drops below 1/10 th the max allowed charging current setting.. I-Max Allowed = 60 A

The capacity monitoring is done by connecting a trumeter digital voltmeter (0-10V) to the battery's connection port, as per Super B manual instructions. I don't know how this works.

As to the system design: It was by the owner builder. I sailed/lived on the boat for three weeks during which time it would always hit about 93% SoC at the end of the day with an absorption voltage of 13.75V. I was told to periodically set that to 13.95 V to keep the State-of-charge accurate. The manual say.s “take it through a complete charge cycle.... This will recalibrate the SoC,” My understanding is it's not good for these batteries to be at 100% SoC but periodically they should be to keep the SoC meter accurate. There is no AC system in this boat, and only a minimu DC circuit, lights, etc.

The solar controller is a “generic” (I assume it came from Aliexpress or similar).

Currently I am living aboard with my wife in the tropics. There are two refrigerators as the main draw with all LED lighting. There is a lot of solar power and it's been sunny every day so it's always gone to the float stage before the end of daylight.

One “problem” I had was at the beginning of my use it was hitting 100% SoC, which I didn't want so I blocked some of the solar panels for a few days so it didn't go into the float stage. When I stopped doing that I could no longer get it up to 90%.

If I had an accurate vote meter I should check the battery values, including SoC output voltage. But the panel meter agrees with that on the charge controller. I believe they are probably more accurate than the crappy multimeter I have.

The previous owner installed an Arduino-based alternator controller. I don't know how to program it but it charges the battery. (Actually there are two motors and controllers.) While it's running the voltage doesn't get higher than the solar controller's absorption set voltage and the batteries aren't warm to the touch.

Does any of this help?

Thanks.

[bradfordharley]

Ignore SOC meters. They calculate a value, and if any of the variables are off (guaranteed) it will be inaccurate. Charge fully in shore power, let rest for a day with zero load and measure voltage. Then reset your SOC meter . There is a lot of info out there in SOC meters and how to tune them.

[wph36]

If you're able, it may be worthwhile to look at the balance state of the cells. If out of wack, this can cause issues at the top and bottom end of the cycle.

[john61ct]

Yes, the only way to know your

bank's ** capacity **

or, how many Ah can it currently store and feed to sources, compared to when it was new ?

is to do well-controlled constant current load test, precisely timed.

The only way to know its SoC other than 100% and 0%, is with a quality coulomb counting SoC meter like Victron BMV, Xantrex LinkPro or Balmar SG200.

Voltage readings have nothing to do with either.

To those not yet purchased an LFP bank, do not do so until getting up to speed, ideally well past these basic concepts.

To the OP, I suggest getting a trusted pro familiar with LFP to come aboard and check things out, teach and advise you.

Maybe everything's fine?

[john61ct]

Originally Posted by Epicurean

Today it did an absorption charge at 13.95 volts. It was set to do this for 300 minutes but only lasted for 45 before switching to the float stage – 13.45V. It says it reverts to the float stage if the absorption time period exceeds the user defined time allowed or if the charging current at that constant voltage drops below 1/10 th the max allowed charging current setting.. I-Max Allowed = 60 A

Personally I think that is way too high and trying to hold way too long.

If the actual charge voltage measured at the bank is hitting 13.95V, stop charging, you are Full.

If you have a proper AH-counting SoC meter, manually reset / tell it at that point that is 100% SoC. Ideally do that each cycle, certainly not less than every few cycles.

Ideally then also turn the charge source off or isolate the LFP from it to carry loads with a lead batt as a buffer. Bring the LFP online before loads require more than the sources are providing.

Or set Float to 13.2V if the above is too much trouble.

[witzgall]

This is why we created the SG200. Once it learns your batteries the inaccuracies of past shunt-based battery monitors are no longer an issue. I remember constantly looking at my circa 2000 meter when cruising each day (many times a day!) and making mental re-calculations based on what I thought the battery SOC "Really" was. I would adjust for PSOC cycling, capacity, etc. All of that is gone now.

Chris

Originally Posted by bradfordharley

Ignore SOC meters. They calculate a value, and if any of the variables are off (guaranteed) it will be inaccurate. Charge fully in shore power, let rest for a day with zero load and measure voltage. Then reset your SOC meter . There is a lot of info out there in SOC meters and how to tune them.

[Epicurean]

I think I can be fully charged, yet that doesn't tell me my capacity. I think I shouldn't be fully charged but still want to have a large capacity.

[CatNewBee]

Originally Posted by Epicurean

I have a problem with the state of charge of mi Super B LifePO4 house batteries. They seem to have lost almost 0.5% of capacity/day over the last two months, although not at a constant rate. The max I can get now is 67% capacity and then they go into a float stage about 2:00 PM, even though the sun is shinning strongly. I charge them through an MPPT solar charger that does a bulk charge then an absorption stage before going to a float stage until they start discharging at night.

Any suggestions?

Thanks

If you realy want to know the capacity, you must perform a capacity test like this:

- Charge the battery to 14.6V ( 3.65V cell voltage ), stay there for at least 30 minutes to allow balancing. Longer is better. You want equal cells. During your charging settings the BMS was very likely unable to balance properly.
- Set your SOC meter to 100%
- Disconnect all charge sources
- Connect a decent load on your inverter and turn it on
- let the battery discharge down to 11.2V under load (2.8V), when disconnect the load, the voltage will probably jump to 11.6...12.0V (2.9...3.0V cell voltage)
- disconnect the load and note the Ah drawn, that is your battery capacity.
- re-charge the battery with your original settings
- when she is going to float, note the Ah charged and calculate the cut-off SOC of your settings, will be around 80..90% very likely. This is the capacity you have with your settings.

Also note chargin cell voltage to SOC depends on temperature.

[Epicurean]

The Super B Manual says there are two ways to fully charge:
1. charge current (C3) until charging at less than 3% of recommended charge current, or
2. set charge current to C3 (52.8A) and end of charge voltage to 14.6 V [seems high to me], then go to float mode (13.7V-13.9V) when charging current drops below 3% of recommended charging current (160A=4.8A)

After doing the latter it should be balanced. The state of charge is done by the BMS sp no idea what happens there.

[CatNewBee]

Originally Posted by Epicurean

The Super B Manual says there are two ways to fully charge:
1. charge current (C3) until charging at less than 3% of recommended charge current, or
2. set charge current to C3 (52.8A) and end of charge voltage to 14.6 V [seems high to me], then go to float mode (13.7V-13.9V) when charging current drops below 3% of recommended charging current (160A=4.8A)

After doing the latter it should be balanced. The state of charge is done by the BMS sp no idea what happens there.

This is the Winston chart LiFePO4, but it is similar to the Super-B, see the voltage to temp dependency. This may explane your observations.

[Delfin]

Originally Posted by CatNewBee

If you realy want to know the capacity, you must perform a capacity test like this:

- Charge the battery to 14.6V ( 3.65V cell voltage ), stay there for at least 30 minutes to allow balancing. Longer is better. You want equal cells. During your charging settings the BMS was very likely unable to balance properly.
- Set your SOC meter to 100%
- Disconnect all charge sources
- Connect a decent load on your inverter and turn it on
- let the battery discharge down to 11.2V under load (2.8V), when disconnect the load, the voltage will probably jump to 11.6...12.0V (2.9...3.0V cell voltage)
- disconnect the load and note the Ah drawn, that is your battery capacity.
- re-charge the battery with your original settings
- when she is going to float, note the Ah charged and calculate the cut-off SOC of your settings, will be around 80..90% very likely. This is the capacity you have with your settings.

Also note chargin cell voltage to SOC depends on temperature.

Cat, isn't the low voltage target that is attainable a function of the BMS used? On mine, the BMS will isolate the battery when it reaches 3.0 volts.