"Floating" LFP batteries

[john61ct]

Finding something useful for excess solar output to do,

should be a completely separate issue from bank care.

There really is no need at all to push more AH in with **any** CV / Absorb stage charging with LFP. Unless precise calibration of 100% SoC is required that cycle.

Choose your target voltage, when it gets there stop. For me, 13.8V / 3.45Vpc is high enough.

Really can't get simpler than that.

If you want to carry loads from the charge source after that point, and feel the need to leave the LFP bank online,

personally I would rather do so at 13.2V / 3.3Vpc than any higher, again no purpose in going higher.

But if truly prioritizing maximum lifespan, No Floating the bank at all is the safest approach.

[newhaul]

Originally Posted by john61ct

Finding something useful for excess solar output to do,

should be a completely separate issue from bank care.

exactly and that is where a holding plate refer system comes into play.

[john61ct]

Or watermaking, cooking e.g. a crockpot, heating water, doing laundry, making ice, lots of options

[nebster]

Originally Posted by john61ct

Finding something useful for excess solar output to do,

should be a completely separate issue from bank care.

Well, that's a powerful word, should. The reality in many systems today is that the challenges of bank care and managing available energy are inextricably linked when a battery provides continuous support for the house power system and when using standard off-the-shelf inverter-chargers and MPPT chargers.

Here's a real world example that I encounter almost every week:

I charge my battery to 3.40Vpc/90% SOC. The next morning, we disconnect from shore power and set out. The battery immediately begins supporting a load that fluctuates from ~0.01C to ~0.20C.

We have 1600W PV available, or ~0.04C. Ideally, this available power would start supporting the load right away. However, the MPPT charger is still in float, where it has been sitting since the charge was completed the day before, and that float voltage is set intentionally low to avoid trickle-over-charging the pack.

It takes some pack drawdown, to say 65%, before my particular float voltage allows the MPPT to start contributing automatically. Only at that point does the battery voltage drop low enough to engage the charger.

Now, you could say, "well, you should get a charger that is smarter and that can be told to start charging at full bore whenever it's needed." And that would be great, but that kind of automation is not readily available at least in my vendor ecosystem... and it's not clear that cobbling something together would be safe and simple and reliable enough anyway.

So that is why I mentioned above that I have an incentive to set higher float values. I am hesitant to set higher float values if it heightens the risk of overcharging or memory-effecting my cells, however!

In my system, these two desires are at odds with each other.

If you want to carry loads from the charge source after that point, and feel the need to leave the LFP bank online,

personally I would rather do so at 13.2V / 3.3Vpc than any higher, again no purpose in going higher.

I happen to float at 3.31Vpc, so you and I have the same gut instinct. But there would definitely be value in going higher... especially when the average load rate is low enough that it takes substantial pack drawdown before the charger can engage. This is a problem more likely to manifest with a larger pack.

[john61ct]

The charge source should be supporting all the loads, no drawing from the bank needed at all unless load watts > input watts.

A low voltage setpoint should not matter at all, yes amps are greater for a given watts level.

Again, a little $20 lead batt can do the buffering, keep the LFP offline until needed.

[nebster]

Originally Posted by john61ct

The charge source should be supporting all the loads, no drawing from the bank needed at all unless load watts > input watts.

A low voltage setpoint should not matter at all, yes amps are greater for a given watts level.

Again, a little $20 lead batt can do the buffering, keep the LFP offline until needed.

You must not have read my example very closely.

If the house battery is offline, what happens when the demand spikes up to higher than the charger can provide? What happens when the cloud passes overhead and the charger can't provide?

In the narrow example of a small system where you're okay taking the battery offline frequently, a lot of things get easier and simpler. But a lot of us have a substantially different situation here.

You should either qualify your statements, or expect that others of us will keep doing it for you. It's misleading to espouse them otherwise, especially on one of the best/only resources on the internet for hobbyists with mobile power systems.

[Delfin]

Originally Posted by nebster

You must not have read my example very closely.

If the house battery is offline, what happens when the demand spikes up to higher than the charger can provide? What happens when the cloud passes overhead and the charger can't provide?

In the narrow example of a small system where you're okay taking the battery offline frequently, a lot of things get easier and simpler. But a lot of us have a substantially different situation here.

You should either qualify your statements, or expect that others of us will keep doing it for you. It's misleading to espouse them otherwise, especially on one of the best/only resources on the internet for hobbyists with mobile power systems.

Most of us just learn to ignore....

[john61ct]

The little $20 batt was an extreme example to make the point. Starter batts, a lead Reserve bank, whatever can be more a more robust option.

Solar has nothing to do with my point, not a significant source IMO for an LFP bank when it refills in an hour or so anytime you run an engine.

My point is just not to let the LFP bank keep sitting there for long periods online with a charge source once it has reached your top end setpoint.

Let the source keep going at higher voltage, keep charging the lead bank(s), carrying loads everything carrying on as normal as if no LFP there.

Bring the LFP online when you know it's needed. I do assume you know before some big load is coming online, to flip a switch and start pulling down the LFP.

Sure if with your setup some big load can power up automatically without you knowing, I suppose then you might need to Float, but that seems an edge case to me.

[newhaul]

:-)I will be using solar only for my Lfp charging and all other power needs so floating the bank will be a reality . I am looking at floating at 13.6 3.4v per cell. Unless there is a good valid reason to go higher or lower .

[tanglewood]

Originally Posted by john61ct

The charge source should be supporting all the loads, no drawing from the bank needed at all unless load watts > input watts.

A low voltage setpoint should not matter at all, yes amps are greater for a given watts level.

Again, a little $20 lead batt can do the buffering, keep the LFP offline until needed.

These may be separate issues to consider, but all interrelated. That’s why we call it a power “system”.

I don’t think there is any “should” in how to approach transferring loads to shore power after LFP is charged. You can disconnect LFP and leave some LA on line as a buffer. Works fine, but involves high current battery switching and control, either manual or automated. And you can switch in other loads when excess power is available, either manually or automatically.

Leaving LFP on line with a standby voltage float works too. There is zero evidence that such an approach negatively impacts battery life, assuming a correctly set float voltage. But the same dependency and assumption exists for charge voltages, absorb times, and all other charger settings. With the “float” approach, there is no high current battery switching, and loads can come and go as they want and the chargers, batteries, and loads will sort it all out themselves. That works fine too.

[Delfin]

Originally Posted by newhaul

:-)I will be using solar only for my Lfp charging and all other power needs so floating the bank will be a reality . I am looking at floating at 13.6 3.4v per cell. Unless there is a good valid reason to go higher or lower .

Lower seems more appropriate. I've noticed that my bank settles into 26.6 volts after fully recharging and sitting for a couple of hours. If you're going to float, and you and Tanglewood have to, then setting that voltage to the resting, no load voltage might be best, or to my mind a bit lower so that the battery sits in the 50 - 60% SoC when in float. Since the discharge voltage curve is so flat I'm not sure how you would find that SoC by voltage alone, so if it were me, I would twiddle with the float setting and compare that to a coulomb counter to try to find the voltage that allows the battery to stay in that sweet storage SoC point over a 24 or 48 hour period.

Nothing much new in this discussion , but interesting.

[john61ct]

Originally Posted by tanglewood

how to approach transferring loads to shore power after LFP is charged.

Again, not talking about any particular type of charge source.

> Leaving LFP on line with a standby voltage float works too.

There is zero evidence this does not shorten lifetime, and I personally see no reason to do so, just like going above 13.8V or pushing into Absorb.

But I agree of course that others can do what they like, as long as they're aware of the issues.

[john61ct]

Originally Posted by Delfin

Lower seems more appropriate.

Yes. Once the "surface charge" peak is removed, 13.2V seems normal, so

if I **had to** Float for some reason

and I wanted to keep SoC high, about to run significant loads from the bank,

I would probably use 13.1V

But when not actively cycling then discharge to say 20% SoC and set the charge source at 12.2V or so,

of course a bit higher if say the bilge pumps may cycle unattended and the non-LFP sources were not robust enough to carry them.

[tanglewood]

Originally Posted by john61ct

> Leaving LFP on line with a standby voltage float works too.

There is zero evidence this does not shorten lifetime,

Sure there is. Equal voltage between two points equal zero current. Zero current is equal to open circuit. So life space is equal to open circuit lifespan.

[newhaul]

Originally Posted by Delfin

Lower seems more appropriate. I've noticed that my bank settles into 26.6 volts after fully recharging and sitting for a couple of hours. If you're going to float, and you and Tanglewood have to, then setting that voltage to the resting, no load voltage might be best, or to my mind a bit lower so that the battery sits in the 50 - 60% SoC when in float. Since the discharge voltage curve is so flat I'm not sure how you would find that SoC by voltage alone, so if it were me, I would twiddle with the float setting and compare that to a coulomb counter to try to find the voltage that allows the battery to stay in that sweet storage SoC point over a 24 or 48 hour period.

Nothing much new in this discussion , but interesting.

I should have been clearer I suppose the "float" voltage will be a daily charge stop and hold untill the sun goes down and I start drawing off the bank.
So it may not qualify strictly as a floating charge since not for long term holding.