Recommended LFP solar charge voltages

[s/v Jedi]

These settings have proven to work extremely well for us. There are some important considerations:

- you charge with solar and are able to fully charge the battery (almost) every day.

- you live aboard and try to not run gensets, motors etc.

This means that you will want to do something with solar production after the battery has been fully charged. When the batteries are full, we convert solar energy into hot water (power the water heater), toilet fresh water (power the watermaker), bread or other baked goods (power the oven) etc. Even when all that would be done and there’s still power available from solar, we run the A/C, removing moisture from the boat and cooling it down a bit.

So here are the important settings:

Absorption voltage: 13.8V for 4S, 27.6V for 8S.
Absorption time: 30 minutes fixed

Float voltage: 13.15V for 4S, 26.3V for 8S.
Float time: until sundown.

The above isn’t suitable for an AC powered charger, alternator etc. It’s just for solar controllers that manage to fully charge the batteries before sundown. It also relies on sundown to stop charging.

So what this means: the battery won’t be full when you reach the absorption voltage, but it will be very close to full after a 30 minute absorption period. Your battery monitor will confirm this and reset to fully charged when configured right (set fully charged voltage below the 13.8 / 27.6V absorption voltage. Check the BMV manual for guidance to get this right.
Just before it resets to fully charged, you can see it’s Coulomb counter (State Of Charge) counting up towards 100% to see how far below full you are.

When the float starts, you will see part of used energy come from the batteries but a large amount is still coming from solar. This is where we start making water, hot water, baking bread etc.
Some days I power my hookah to clean the bottom etc. Use the available power that you can’t store in battery.

[PaulCrawhorn]

13.8V is a good definition of Full as a termination voltage, but the actual SoC at that point will depend on the current level used to get there.

A lower current gets to a higher SoC, for a given termination V.

There is no useful reason to strive for Full with LFP. Even just getting "close to Full" is harmful to longevity, especially if the cells will sit in that state for more than a few hours.

Thus there is no reason to hold Absorb at all in normal cycling charge profiles.

If loads are then drawing the bank down then maintaining that low Float voltage will be healthy for the cells, but then why go to Full in the first place?

If setting your BM to an objective definition of 100% is important to you, then OK, Absorb stage is needed for that to be somewhat standardised, but a trailing amps cutoff more so.

The above points are just general raising awareness for systems design, for those prioritizing longevity of the banks, I am not trying to dispute the main overall points made above.

[s/v Jedi]

Originally Posted by PaulCrawhorn

13.8V is a good definition of Full as a termination voltage, but the actual SoC at that point will depend on the current level used to get there.

A lower current gets to a higher SoC, for a given termination V.

There is no useful reason to strive for Full with LFP. Even just getting "close to Full" is harmful to longevity, especially if the cells will sit in that state for more than a few hours.

Thus there is no reason to hold Absorb at all in normal cycling charge profiles.

If loads are then drawing the bank down then maintaining that low Float voltage will be healthy for the cells, but then why go to Full in the first place?

If setting your BM to an objective definition of 100% is important to you, then OK, Absorb stage is needed for that to be somewhat standardised, but a trailing amps cutoff more so.

The above points are just general raising awareness for systems design, for those prioritizing longevity of the banks, I am not trying to dispute the main overall points made above.

Exactly, termination voltage in combination with charge rate determines the SOC when that voltage has been reached and this is different for every setup. With our solar, the numbers I posted work great.

I don’t know much about other battery monitors, but the Victron Smart BMV is a Coulomb counter and it requires the resets at 100% SOC for calibration. When you don’t do this, the error will keep compounding until the readout has become inaccurate.

Going to full charge before doing the lower float setting to retain SOC while the solar is still delivering is done to ensure battery SOC at sundown and BM accuracy.

[tanglewood]

I think absorb is needed for some amount of time by a lot of BMSes to do balancing. They only balance when the voltage is above a certain level.

[s/v Jedi]

Originally Posted by tanglewood

I think absorb is needed for some amount of time by a lot of BMSes to do balancing. They only balance when the voltage is above a certain level.

Yes, unfortunately these BMS’s do that. This is another point I think needs to be done very differently: the BMS should determine if balancing is required and then turn on an external balancer that can work at whatever level the BMS wants balancing being performed.
A simple $40 active balancer from Amazon can do that job.

[Pete7]

If the float voltage drops to 3.15v per cell what percentage of capacity is this? depends on the cell, but about 70% perhaps.

If you have and I have seen the pictures a large bank, it may not matter. For those with smaller banks, it might be better to opt for a slightly higher float in the 80-90% range, so which gives greater available power. Needs to be chosen with care so you don't end up floating at 100% all the time.

We are currently playing around with a 13.3v setting but that's not a recommendation and we may change it.

The other point of the float is to maintain the LFP but also supply all the domestics predominantly in the afternoons so batteries nearly full when the sun is much lower in the evening. Then really start to draw on the LFP.

[s/v Jedi]

Originally Posted by Pete7

If the float voltage drops to 3.15v per cell what percentage of capacity is this? depends on the cell, but about 70% perhaps.

If you have and I have seen the pictures a large bank, it may not matter. For those with smaller banks, it might be better to opt for a slightly higher float in the 80-90% range, so which gives greater available power. Needs to be chosen with care so you don't end up floating at 100% all the time.

We are currently playing around with a 13.3v setting but that's not a recommendation and we may change it.

The other point of the float is to maintain the LFP but also supply all the domestics predominantly in the afternoons so batteries nearly full when the sun is much lower in the evening. Then really start to draw on the LFP.

My float setting isn’t 3.15V per cell… it is 13.15 for 4 cells, i.e. almost 3.29V per cell. I have not seen less than 87% SOC before going to sleep so that’s with all electric cooking after sunset as well as using much more energy in late afternoon than what solar produces, like for the watermaker, a hookah pump, oven or even A/C. Like you write: use all the solar array can give and after batteries are full, store it in the form of fresh water, hot water, baked bread, scraped underwater hull, cool and dry interior etc.

[Pete7]

Oops, well spotted, my mistake.

Looks like we should call on you late afternoon then for afternoon tea

[PaulCrawhorn]

Going to no-matter-how very high / long definition of 100% (not advised)

once surface charge is dissipated, isolated and at rest LFP will sit between 3.32 and 3.35

That is the most objective definition of 100% there is, and 1000 charge profile variations will get you there.

No point trying to squeeze any more mAh in there, just causing stress without actually adding any useable stored energy.

And sitting there, or anywhere near 3.3V is shortening lifespan.

If you think your poorly designed balancing gear, or your BM "requires" doing stuff that reduces lifecycles, personally I would look at replacing that circuitry. Personally.

From that 100% SoC down to your definition of 0% is a pretty shallow but steady voltage curve, at least at a low and steady C-rate, better to use a cutoff of 3.15 rather than 3.05Vpc.

With high C-rate loads, that cutoff needs to be lower to account for the voltage sag, bounces back up once the load is off.

Leaving a higher capacity unused on average at this bottom end can double or triple cycle lifespan, so it's worth buying a bit more Ah capacity than you usually need. Certainly better than hesitating to draw down after a charge cycle.

Or fire up your fossil fueled charge source a bit earlier, if you're going to have to anyway before solar starts up tomorrow, those cost fuel and wear & tear though, so that's a balancing act...

Those not concerned about getting extra thousands of cycles out of their LFP bank feel free to ignore these finer points with good cells it's pretty hard to hit less than 2-3000 cycles so long as your setup prevents disastrous "unexpected events"