LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[john61ct]

Originally Posted by JmanC

By the looks 13.1V is better as it keeps it in the 30-40% region which I hear is better than 50-60.

If you believe letting the LFP bank sit with constantly active charge input is fine, there will be little difference between 3.28Vpc and 3.30.

Even 12.4V (3.1Vpc) should be fine, maybe even better.

IOW once you're well away from the top SoC area, precision in voltage/SoC is not important, other than of course taking no chances of dropping below say 3.0Vpc.

_____
But for those who prefer to define their storage protocol as isolated from all possible risk factors - including potential failures and vampire loads from "protective" circuitry,

while shore power (or other constant/reliable input source) is carrying all the load devices anyway,

a lead Reserve, or any old cheap battery, Starter for that matter, can easily act as the DC buffer rather than putting the LFP bank at risk.

[john61ct]

Originally Posted by JmanC

I hear higher voltages like 14V+ mixed with constant low & partial charging mean possible damage overtime?

No, in fact if you **know** you won't be getting near Full SoC, higher setpoints are safer (as long as not exceeding the max spec) than when going to high SoC.

In theory, a charging algorithm that pushed higher voltage and currents in the early low-SoC stage, then scaled both back when approaching Full could speed up the charge cycle without reducing longevity. But I haven't come across a marine charger implementing this.

Alternatively, if you are referencing the possibility of a temporary "memory effect",

the chances of that resulting in permanent capacity loss is remote, as long as the PSOC stop-charge point varies, and especially

if you can push through to the maker-spec 100% every few weeks or so.

> You should ideally land in the 90-95% soc

When you are cycling, need to soon be servicing loads, otherwise sitting at lower SoC is better.

[CatNewBee]

Originally Posted by rgleason

Mainesail seems to manage their battery discharge and charging manually.

"..For our particular design use I don't really need a float setting. When we are done charging, charging is simply off. ... otherwise we don't turn on charge sources until we've dropped to about 20% SoC.

Catnewbee:

"I do use 13.5V as float setting for solar, it translates to 3.375V cell voltage during the day."

"While cruising, absorption is set to 14.4V, we do use a lot of power for the water maker, induction cooking, ...etc ...... Even with 1650Wp solar we get to absorption only a few days of the month.."

Catnewbee is this basically while you are on board using power?

Yes, this is the setting while cruising and using the power.
Last year we left the vessel in the marina with no shore power connected, all switched off except bilge pumps an AIS and solar. It kept the start batteries happy, at reduced voltages, Absorption was set to 13.8V and float to 13.4V.

After the winter the BMS showed 80% SOC, I don't know it it was good or bad for the batteries, they are working as they did before. This year we will leave only for some weeks and will stay on board during the winter, so the batteries will be cycled to some extent along with shore power for charging and power assist for heating, cooking, baking etc. Marina has only 220V 16A outlets, so the battery will step in when necessary.

[CatNewBee]

Originally Posted by mikedefieslife

I thought that you don't use absorption at all for LFP? Pretty sure most things I've seen say set the absorption time to 0 on Victron products.

Absorption is a set point, that terminates BULK at this point and keeps the voltage for a while (ABSORPTION TIME). Originally I had it set to 5 minutes, but now after seeing I don't regulary reach full charge, I increased the time to one hour to allow some balancing of the cells. I will reduce the value as soon as I am happy with the balance results.

[john61ct]

Yes, depending on how the balancing circuitry works,

that may require holding the end of charge stage (above the start-balancing voltage setpoint) for a long time

in order to complete the balancing process.

How long depends on the balancing current available from that circuitry, and how unbalanced the cells have gotten.

All these considerations are a completely separate issue from what is healthy for the LFP cells themselves.

Many find their (always attached) protective BMS does such a poor job balancing, they disable that function,

using dedicated better-designed balancing gear (or just manually with an adjustable PSU)

to do the job periodically as needed as a separate maintenance routine.

With good quality cells well cared for, as long as they remain relatively healthy,

this may only need doing after many months of normal cycling, or even years.

[CatNewBee]

... In theory.

[CatNewBee]

...

Balancers work with small currents, about 1..2A. A 1000Ah cell, that runs by 1% out of balance in PSOC cycles needs 5..10h to become balanced again, not achievable without some absorption time.

[john61ct]

Yes, that is exactly my point, most protective BMS balance in that manner.

1000Ah cells are an extreme example, in that scenario one would think it behooves the owner to not let them get so far out of balance.

But in any case, plenty of other balancer types work at current levels higher than an amp, usually (but not always) the current increases with the voltage delta between cells.

And many can do their balancing - at least start the process - at voltages well below Full.

In fact. some are designed to run 24*7, at **all** SoC levels.

And many owners just balance (when needed) manually, using a simple adjustable DC converter or PSU.

As would I think always be the case for those choosing bottom balancing.

[mikedefieslife]

Originally Posted by CatNewBee

...

Balancers work with small currents, about 1..2A. A 1000Ah cell, that runs by 1% out of balance in PSOC cycles needs 5..10h to become balanced again, not achievable without some absorption time.

I thought that active balances work without a charge current. They bring the higher voltage cells down and the lower voltage cells up.

I bought an active balancer board to play with, but it seems it's missing in the post

[john61ct]

Originally Posted by mikedefieslife

I thought that active balances work without a charge current.

Yes, many different types do.

The type you're referring to is aka "capacitive balancing BMS" or "shuttling" or "active*transfer"

Similar to resistive bleeding type, but instead of bleeding off the excess energy from high cells via heat, they transfer it to other cells in the pack.

[CatNewBee]

Originally Posted by mikedefieslife

I thought that active balances work without a charge current. They bring the higher voltage cells down and the lower voltage cells up.

I bought an active balancer board to play with, but it seems it's missing in the post

It makes sense to balance only if there is a charge current, otherwise you will discharge the battery over time, it also makes sense to balance only at the shoulders, the voltage difference is more significant to make the right decision.

I know, John is very knowledgeable in the theory and publications and very focussed on longevity instead of usage, real technical solutions are always compromises and you must make the right ones that work for you.

I do have an active BMS btw, and it is set to kick in early, but well above resting voltage in conjunction with charge current at 3.375V, balancing without charge current happens in my setup only above 3.382V, so in theory an overcharged cell would be balanced down at float, but as i wrote at PSOC charging without hitting absorption and float, the cells run out of balance over time.

And NO, it is not a practical solution to remove the cell isolation on the bars and do manual cell balancing with a lab power supply or other adventurous setups. This is the wrong approach, the battery should remain sealed (all poles and connections are enclosed in liquid rubber for safety and corrosion protection) and operational, and the BMS has to be empowered to do his job, and it will when you let it work as supposed.

We are talking about a liveaboard bord battery and not about bench testing and fiddeling around in the lab.

The solution at the moment is float at 13.5V on shore power, the cell difference is after 48h now only 8mV between highest and lowest cell, so almost in balance again, the BMS has done a great job and I am happy with the result.

[CatNewBee]

Image could not be displayed, who knows why...

[john61ct]

Of course balancing circuitry should not be allowed to discharge the bank when not needed.

This is not at all an issue when only used for a few hours per year, connect for that time, isolate otherwise.

If used every charge cycle, then usually left in place all the time, but the ideal then is the balancer works within your preferred charge profile.

It is not a goal to top-balance well below the shoulder, but if the balance rate is too low, you need to **start** the process at a lower voltage otherwise the bank is sitting at full charge for too long.

And yes, only a concern if you care about extending longevity.

Manual balancing does **not** require breaking apart the bank into individual cells, nor any lab bench setup, if you design your bank with that in mind.

Obviously, if you get OceanPlanet, Victron or Mastervolt to just install a packaged system for you, you will use the balancing gear and procedures they have designed.

Plenty of advantages going that way if you can/want to pay those kinds of prices.

[JmanC]

I'm having trouble with my BMV 712. It's soc gauge doesn't stay inline with my dropin's builtin bluetooth gauge & looks to be inaccurate. By the time I discharge down to 20% soc (according to bluetooth soc) with a 0.25C discharge load the BMV is reading only 16% SoC.
The Bluetooth app soc always turns to 0% at exactly near 12V under load regardless of loads, C/20 or even .3C
Before this discharge test the BMV was manual re-sync at 100% full.

Apon recharging, the BMV soc doesn't lift from 0% until the bluetooth gauge is 7%.

The battery is a 125AH battery & bluetooth app shows about 134AH total capacity which is exactly what the BMV AH consumed reads when 12V under load is reached regardless of loads. The BMV also has it's capacity set to 134AH.

So what do I need to do to get the BMV inline with the bluetooth soc gauge?
Use another capacity? If so what?
Change some settings?
Main ones are:
peukert is 1.05
charge eff factor is 99%

[CatNewBee]

Originally Posted by JmanC

I'm having trouble with my BMV 712. It's soc gauge doesn't stay inline with my dropin's builtin bluetooth gauge & looks to be inaccurate. By the time I discharge down to 20% soc (according to bluetooth soc) with a 0.25C discharge load the BMV is reading only 16% SoC.
The Bluetooth app soc always turns to 0% at exactly near 12V under load regardless of loads, C/20 or even .3C
Before this discharge test the BMV was manual re-sync at 100% full.

Apon recharging, the BMV soc doesn't lift from 0% until the bluetooth gauge is 7%.

The battery is a 125AH battery & bluetooth app shows about 134AH total capacity which is exactly what the BMV AH consumed reads when 12V under load is reached regardless of loads. The BMV also has it's capacity set to 134AH.

So what do I need to do to get the BMV inline with the bluetooth soc gauge?
Use another capacity? If so what?
Change some settings?
Main ones are:
peukert is 1.05
charge eff factor is 99%

You could reduce / tweak the Peukert, the charge efficiency is only a reduction while charging, means the BMV shunt must be passed by 100Ah to show 99Ah charged, 1Ah losses by heat (). The Peukert is pretty good explained in the manual. It boils down to capacity (SOC) loss based on Ah drawn and Amps flowing. Lead acid batteries have less capacity on high current rates. You can try a Peukert of 1.01

You also could play with the shunt settings if your current readings vary from the BT and BMV, but usually the BMV are well calibrated ex works, so I would suspect the error in the BT interface / BMS.

More important are the voltage settings for absorption and absorption time, they are responsible for synching to 100%