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

[Maine Sail]

Originally Posted by CharlieJ

Diugo-
I thought that there were more manufacturers with a float voltage spec but it appears that CALB, Winston and Sinoply do not. Forzatec (cells that I am using on 2 x 1000Ahr packs on a sailing cat) does provide a 3.4VPC float voltage spec: FORZATEC | LifePo4 batteries manufacturer | Lithium ion batteries manufacturer

That said, my experience with a the 4P4S battery pack I have in the shop is that the CALBs float very well @ 3.35VPC with < 1 A flow into the battery pack. As power boats routinely float there LA batteries for extended periods under analogous conditions (alternator voltage very close to battery bank voltage for low current flow), I am not sure I understand what the proponents of not floating LFPs are concerned about.

Charlie,

I know the CALB cells are a slightly lower voltage than Winston but for my pack, at 13.4V, no current is flowing as the bank will rest higher than, for a good period of time, even with a 1A+/- parasitic load ..

If I discharge and recharge to only 13.4V the current continues to flow for a while because the bank is still "charging" all be it very, very slowly. I would not however consider this a "float" if I had not gone above that voltage first, and would still consider it "charging"...

I plan to charge to 13.8V then step down to a "float" that is below resting which will essentially stop charging. Necessary at below the upper knee? Not IMHO, but I still plan to step my charge sources down to below resting after I charge to my set point..

I am not going to charge or push into the upper knee. Even at my 13.8V the bank will take far more current than our vessel can provide so unlike LA batts this lower charge voltage is not really slowing my charging.

With now over 45 varying cycles on my pack, and none except the first balancing cycle have been into the upper knee, the cells have remained in 100% perfect balance charging to either 13.6V, 13.8V or 14.0V. I've experimented with everything from 13.4V to 14.2V as my top charge point.

I am planning on using 13.8V as my upper end of charge for the Winston cells. This gives me tons of Ah capacity and keeps me well out of the upper knee, something I just feel more comfortable with..

As it looks now I will probably re-balance them once per year manually in parallel configuration and by pushing the voltage to 3.8VPC as I did initially. If I were to ever need a mid season "balance" I can simply bring my power supply down, adjust the current and voltage, and push into the shunting range of the BMS. As it looks now this will not be necessary...

My BMS is there only as an emergency back up to my charge & load system. I simply don't want to push to cell level shunting on every cycle as I have found, thus far, it is not going to be necessary to do so.

[diugo]

My concern is that floating above the resting voltage---even under near zero charge current---subjects the cell to a constant polarizing electric field, which could encourage some of the intercalated lithium to permanently bind to the graphite in the anode, resulting in premature loss of capacity.

This effect might be less pronounced in cells that are deeply cycled on a daily basis---and more pronounced in those that will be lightly cycled (like mine).

It's definitely controversial, as life sure would be easier if we could simply use a ubiquitous LA charger on LFPs and just set it to float at a convenient voltage close to full charge. Unfortunately, there is ample evidence that "lithium ion" batteries are degraded by floating. Some have claimed that the LiFePO4 chemistry is somehow immune from this. I'd just rather not use my expensive cells to find out.

[Maine Sail]

Originally Posted by diugo

My concern is that floating above the resting voltage---even under near zero charge current---subjects the cell to a constant polarizing electric field, which could encourage some of the intercalated lithium to permanently bind to the graphite in the anode, resulting in premature loss of capacity.

This effect might be less pronounced in cells that are deeply cycled on a daily basis---and more pronounced in those that will be lightly cycled (like mine).

It's definitely controversial, as life sure would be easier if we could simply use a ubiquitous LA charger on LFPs and just set it to float at a convenient voltage close to full charge. Unfortunately, there is ample evidence that "lithium ion" batteries are degraded by floating. Some have claimed that the LiFePO4 chemistry is somehow immune from this. I'd just rather not use my expensive cells to find out.

As a point of reference Genasun, who understand this quite well, allow their banks to be charged at 14.2V. Their Lithium solar controller charges to 14.2V and just stays pumping our current as long as there is sun and the batteries will take it. They use CALB cells and stand behind the product even at the suggested 14.2V.....

[diugo]

Very interesting find! Digging a bit deeper, 14.2V is considerably higher than the 13.6V (3.4VPC) float voltage recommended in the official Calb manual at http://www.manzanitamicro.com/downlo...load=80%3Acalb

[hellosailor]

Charlie-
" Several of the LFP cell manufacturers lists a float voltage on their spec sheet."
Perhaps that means the voltage shown is the optimum volage for float charging? As opposed to recommending float charging at all. Two very different things.

Look at a modern laptop computer, and you'll see two different charging regimens under the power management options. One allows for keeping the battery at 100% capacity all the time, for those users who want to know they can take it away with the longest possible charge. And a second protocol, which may let the battery drop to 95% before it recharges at all, because that promotes longer battery life overall.

Publishing a spec, or listing an optimum spec for a purpose, does not necessarily endorse that type of operation versus another one.

That's all it would take to explain the contradiction, with or without translation errors.

[deckofficer]

Some very interesting discussion. The EV gang charges at 3.7 per cell (14.8 for 12 volts) for the fastest full charge. For my kayak propulsion bank I do the same. With house batteries it is a different animal and I tend to lean towards Maine Sail's methods for when I do a house bank.

"I plan to charge to 13.8V then step down to a "float" that is below resting which will essentially stop charging."

[diugo]

Looking more closely at the Genasun products, what they seem to refer to as "float" voltage is actually just "CV voltage" in the specs---as in CC/CV. In other words, they are simply holding at 14.2V until the current tapers to a particular value, then terminating the charge. Not floating at 14.2V indefinitely.

Nevertheless, the 3.4VPC float spec in the CALB document remains intriguing. I'll ruminate on it a bit more before presenting my thoughts.

[cgegner]

Serial numbers don't assume the are identified early in manufacturing.

[Maine Sail]

Originally Posted by diugo

Looking more closely at the Genasun products, what they seem to refer to as "float" voltage is actually just "CV voltage" in the specs---as in CC/CV. In other words, they are simply holding at 14.2V until the current tapers to a particular value, then terminating the charge. Not floating at 14.2V indefinitely.

Nevertheless, the 3.4VPC float spec in the CALB document remains intriguing. I'll ruminate on it a bit more before presenting my thoughts.

Nothing in the Genasun system "terminates" the charge unless you go over voltage on a cell and the BMS does a HVC. They are simply charging to 14.2V and it will say there as long as the source keeps 14.2 present. The current will eventually taper to virtually nothing but I am unaware of anything in the system which "terminates" the charge.

I was just working on a large Genasun install three weeks ago and it has solar (Genasun controllers), Victron I/C and a Balmar regulator... All charge sources were programmed for 14.2V. That is just how Genasun designed the system to work...

[diugo]

Genasun describes its lithium charge procedure simply as "CC/CV". I think most here assume that "CV" means "apply constant voltage until X."

Genasun sells a solar charge controller for LiCoO2; does it really subject a volatile cobalt-based lithium-ion battery to a constant 16.7V indefinitely??? I somehow doubt it.

Most lithium chargers employ a C/20 or similar cutoff scheme, but there are other methods. For example, a periodic topping charge: disconnecting the voltage source as soon as the CV voltage is attained, and then reconnecting it only whenever the battery voltage falls below a delta.

This is much different than floating, where the voltage source is always connected. Genasun clearly differentiates the two terms in its spec sheets, taking pains to use "float" only for its LA chargers and "CV" only for its lithium ones.

IMHO this suggests Genasun does not truly float its lithium batteries.

[daddle]

Originally Posted by CharlieJ

daddle-I, and I am sure others, appreciate your effort to find the tech paper.

As an aside, I also agree that the LFP datasheets leave much to be desired. ...

This is my favorite paper on the subject of LFP cell durability. It can be a bit academic in places but there are good simple narratives of all the important points.

My assertion above is from Section 4.4.3 on pages 38-39.

If yacht users follow T1 Terry's practice of cycling only between the "knee" regions of the discharge curves then most of the issue can be avoided. Of course at the cost of the system's glorious and over-hyped performance being somewhat diminished.

The discussion on CF about "float" may be because of terminology coming over from LA technology. In a LA cell "float" generally means a long-term low charge current. With LA this promotes long life. Posters here are correct that a low "float" voltage in a LFP system is much the same as stopping the charge because the charging current into a LFP cell will be zero.

In a typical cruising (active) yacht application the idea of "float" is not so useful in practice as there is no steady state. Large discharge loads are frequent as well as irregular charging (solar shading, wind lulls, periods of motoring). The cells switch between charge and discharge frequently.

It will promote the LFP's long term health if the charge sources are always at the ready to supply power to the yacht's loads. That way the power does not need to come from the cells which reduces the cycle count or "power processed". It would be a pity if the solar regulator did not supply the refrigerator power because it's "smart" rules determined that the LFP cells were fully charged .... for example.

[Lagoon4us]

Australian GSL MPPT built soley for Lithiums settings are as follows:-
1) HIGH CAPACITY 14.8 >
2) DEFAULT/FACTORY 14.6 > SATURATION COMES IN AT 13.9V FOR SETTINGS 1,2,3
3) LONG LIFE 14.3 >
4) MAINTENANCE 14.1.
__________________

[FlyingCloud1937]

Originally Posted by diugo

Genasun describes its lithium charge procedure simply as "CC/CV". I think most here assume that "CV" means "apply constant voltage until X."

Genasun sells a solar charge controller for LiCoO2; does it really subject a volatile cobalt-based lithium-ion battery to a constant 16.7V indefinitely??? I somehow doubt it.

Most lithium chargers employ a C/20 or similar cutoff scheme, but there are other methods. For example, a periodic topping charge: disconnecting the voltage source as soon as the CV voltage is attained, and then reconnecting it only whenever the battery voltage falls below a delta.

This is much different than floating, where the voltage source is always connected. Genasun clearly differentiates the two terms in its spec sheets, taking pains to use "float" only for its LA chargers and "CV" only for its lithium ones.

IMHO this suggests Genasun does not truly float its lithium batteries.

That is my exact interpretation of Float Charging. All US and Canada remote cell/micro sites use what is called a Cycle Charge/Float logarithm.
aka standby.

All lead acid bats suffer from a constant voltage charge, it causes positive grid plat corrosion. Doesn't matter whether its FLA, SLA, or Gel. I'm also understanding that Li-bats have the same issue.

That's why most deep cycle AGM's suffer low terminal life, at least the ones taht are held at a constant CV

Magnum, and Outback have a battery saving logarithm that cuts CV(off) on float until the bat drops to a pre-set voltage, then it starts back to CV again. This is what a true float voltage charge is.

Charge Cycle means discharge/re-charge until float V, then drop the charger until V drops reaches X, then CV for pre-set again, this equals bats are in stand-by until the call to duty.

Solar charging will most likely fill this need, as the sun goes down every evening so CV quits when the SP's do.

I'm going to bet that these bats will suffer much like AGM's if they are left to CV on a shore charge.

Lloyd

[ebaugh]

Originally Posted by FlyingCloud1937

Magnum, and Outback have a battery saving logarithm that cuts CV(off) on float until the bat drops to a pre-set voltage, then it starts back to CV again. This is what a true float voltage charge is.

Lloyd

Yes Magnum does. It drops back to a temperature adjusted 12.8V, then recharges to float voltage. And without the latest hardware and firmware, it can't be defeated even when using custom battery settings. This is a huge problem when plugged in at a marina.

[ebaugh]

Originally Posted by daddle

This is my favorite paper on the subject of LFP cell durability. It can be a bit academic in places but there are good simple narratives of all the important points.

Thanks for posting this! The references section is also extensive and the two or three I searched for so far are out there for free downloading somewhere.