How bad is it to charge LiFePO4 only via constant voltage?

[sn34ky]

I was wondering how bad would it actually be to charge LiFePO4 batteries only at constant voltage (3.6V each cell) and low current (0.5C at most). Does anyone know how much would that decrease battery life, or if it is dangerous?

 

[John Frum]

I was wondering how bad would it actually be to charge LiFePO4 batteries only at constant voltage (3.6V each cell) and low current (0.5C at most). Does anyone know how much would that decrease battery life, or if it is dangerous?

For most of the flavors of LFP I'm aware of .5c is not low current.
In fact its the maximum that you should use.
.2c to .33c preferred.
If you only charge constant current the cells will very likely get increasingly un-balanced with ever cycle.

 

[Ampster]

was wondering how bad would it actually be to charge LiFePO4 batteries only at constant voltage (3.6V each cell) and low current (0.5C at most). Does anyone know how much would that decrease battery life, or if it is dangerous?

Actually when you apply a constant voltage to a cell the voltage at the interface drops to the cell voltage. People charge to 3.6 V all the time. I have grey market cells and chose to charge to 3.45 V per cell just to increase my odds. I can't quantify if any there is any life expectancy increase but I can quantify that there is not much capacity gained from 3.45 V to 3.60.

 

[John Frum]

For most of the flavors of LFP I'm use to .5c is not low current.
In fact its the maximum that you should use.
.2c to .33c preferred.
If you only charge constant current the cells will very likely get increasingly un-balanced with ever cycle.

UPDATE: I see you have said constant voltage.
It makes no sense to charge with only constant voltage.
Constant voltage would mean that battery voltage was close enough to the charger voltage that the charger could not deliver its configured/rated amperage.

 

[Ampster]

If you only charge constant current the cells will very likely get increasingly un-balanced with ever cycle.

To clarify, what would cause that is the lack of time at Constant Voltage stage for balancing of the BMS to start or have any effect.

 

[John Frum]

Doco attached.

 

[stienman]

This 2021 paper presents an overview of the research available on float charging lithium ion batteries:

Overview of research on float charging for lithium-ion batteries

You may need to translate the article - it's in Chinese.

I didn't dive into it deeply, however it does cover a lot of ground about the consequences of float charging, from capacity loss to effects in battery packs vs individual cells to safety concerns.

The short answer is that you can expect capacity loss (which is highly dependent on temperature, charging voltage, and charging current), and gas production (internal cell pressure increasing). How much and whether it's particularly dangerous or not is variable based on the exact cell chemistry, construction, and usage.

I would recommend following the cell/battery manufacturer's instructions, which probably says no float charging, and if you find yourself in the position where you have to go outside the manufacturer's recommendations then do thorough testing - including safety testing - before deploying a float charged cell/battery.

 

[JoeHam]

I can quantify that there is not much capacity gained from 3.45 V to 3.60.

Exactly.

If you’re going to charge with CV I would suggest 3.4 to 3.45 V.

I rarely go above 0.2 C but that’s extra conservative.

 

[Substrate]

Keep in mind that nearly all recommendations / papers for operation, while good, are assuming 24/7 ac supply use, and not solar, where the universal power switch is turned off at night.

In other words, let's say you reach your CV point and hold, but there's only 30 minutes of daylight left.

This is far less damaging than obviously using an ac-powered charge source that never shuts off.

I'm just saying that sometimes the application recommendations don't *always* apply *completely* to solar use. You'll have to be the judge and use some common-sense. In particular, identify your application - are you mostly standby / weekend warrior, or daily-cyclic?

Here, the values and concerns may change a little bit.

 

[Goboatingnow]

True constant voltage means that the supply will maintain the voltage under all load conditions. Ie it will have infinite current available

Such a system would destroy your lithium battery.

Real life constant voltage have power limits. Upon reaching that limit they “ may” drop to constant current. This is how most lithium chargers work anyway. A CC phase where the charger delivers all the current it can followed by a CV where the battery current tapers down to zero.

For optimum Li life once the current tapers to close zero. The impressed voltage should be removed. Do not float Li. Even dropping to a lower voltage is bad.

 

[Substrate]

The fun is in the details however.

True constant voltage means that the supply will maintain the voltage under all load conditions. Ie it will have infinite current available

Until the sun goes down. If it reach full charge at breakfast time, not so good. If you reach full charge late in the afternoon, and 10 minutes later the sun goes down throwing the big universal master switch, not so bad.

Thing is, generalized statments cannot be relied upon as gospel because they lack details. Did you know that if I hold my CV at 3.4v / cell or less, I'll never reach full charge, and current drops to zero without harm - it is lower than what LFP will absorb at, and most importantly because of this, the *secondary reactions* of holding it here is not degrading.

For optimum Li life once the current tapers to close zero. The impressed voltage should be removed. Do not float Li. Even dropping to a lower voltage is bad.

Charging at voltages higher than LFP's natural absorb voltage of about 3.4v, and holding it there until zero amps is bad. There is a tail-current value at which you stop. If you don't, you anger the little secondary-reaction gremlins which add up over time.

Typically, this tail-current value is about .05C of the batt capacity (5%) upon which you should stop.

If you charge LFP with say a 14.0 to 14.6v CV, or even if you just reach it and stop, if you fall back to float it isn't harmful. WHY?

Most float values are 3.4v per cell - the typical knife edge that you have to exceed to fully charge. (more like 3.45v) For a nominal 12v battery, that would be 13.6v. 13.6 / 4 cells = 3.4v cell.

Part of the reason you drop to float, is that in many rv / vehicular applications, this can act as a parasitic-load catch so that you don't walk out to the RV and find you have tripped your bms low-voltage trigger.

What I'm saying is that generalized sweeping statements need to be taken with a grain of salt depending on your application. Applying lab-perfect operations may make your system impractical to use in operation. Heh, there are those that treat their systems so perfectly, they get nearly no use out of them, and outlive their chemical viability.

 

[Ampster]

What I'm saying is that generalized sweeping statements need to be taken with a grain of salt depending on your application.

I agree and I follow your recommendations. I also agree with @Goboatingnow that if someone does not follow what you are saying to the letter, then they could create a problem without knowing why.

 

[HighDesertOffgrid]

Applying lab-perfect operations may make your system impractical to use in operation. Heh, there are those that treat their systems so perfectly, they get nearly no use out of them, and outlive their chemical viability.

I am trying to wrap my head around real world application of this. My system is cycling all day every day with water heater elements or ac compressors kicking on and off sometimes drawing several kilowatts. It seems better to establish a float voltage than to risk near dead batteries toward sundown.

 

[Ampster]

It seems better to establish a float voltage than to risk near dead batteries toward sundown.

Or at least at sundown have your batteries at full resting voltage when the sun goes down. That is better than entering darkness with only 80 percent capacity.

 

[HighDesertOffgrid]

Or at least at sundown have your batteries at full resting voltage when the sun goes down

I don't know how that would even be possible with an off grid system like mine without running a gas generator every evening to make up whatever amps would be lacking (which would always change).

 

[Substrate]

People charge to 3.6 V all the time. I have grey market cells and chose to charge to 3.45 V per cell just to increase my odds. I can't quantify if any there is any life expectancy increase but I can quantify that there is not much capacity gained from 3.45 V to 3.60.

No not much capacity difference, but at 3.45v, you are just slightly above the LFP's absorb "knife-edge" where given enough time (like a week or more) you will be actually full capacity if you do a discharge capacity test. At 3.6v CV, you may reach a full charge (capacity wise) in 10 minutes.

An example from a different application:
Site chooses to use LFP for a ups application. Manager thinks that 3.45v is really low in CV and won't fully recharge. Comes back a few weeks later, does a capacity test, and holy cow! Full capacity! Cells sitting in a ups standby environment at full capacity is not good.

It's a *time* issue. Given enough time, a CV of 3.45 will *eventually* fully charge the cell. But oh boy, that dinky little current. Eventually, full capacity. At 3.4v or below, LFP will not chemically allow for an absorb to full. Sure it will absorb to zero given enough time, but that is OK, because chemically, it will never achieve a full capacity.

In the above example, had the UPS manager dropped his CV to 3.4v, maybe 3.35v, he could float at that and never reach full capacity. Sure he wastes some of the capacity he paid for, but he wisely planned for de-rating his cells operating under the LFP absorb "knife edge".

Too many words, but if you've got the time, it's fun to test. Don't use your best cells for the 3.45v CV two-week to absorb to full capacity test.

 

[Ampster]

I don't know how that would even be possible with an off grid system like mine without running a gas generator every evening to make up whatever amps would be lacking (which would always change).

My cells are full by Noon and I charge to 3.45 V per cell. My loads are covered by solar until sundown so I enter darkness with a virtually full pack. Sometimes in winter it does not happen every day but it does not matter with Lithium. I am not off grid but my Hybrid inverter is in self consumption mode which is effectively the same. If you cannot get a full charge by afternoon during most months you do not have enough solar.

 

[HighDesertOffgrid]

My cells are full by Noon and I charge to 3.45 V per cell. My loads are covered by solar until sundown so I enter darkness with a virtually full pack. Sometimes in winter it does not happen every day but it does not matter with Lithium. I am not off grid but my Hybrid inverter is in self consumption mode which is effectively the same. If you cannot get a full charge by afternoon during most months you do not have enough solar.

I'm usually charged by noon as well. This time of year my energy production starts to fall off around 4pm though. I am absorbing/floating at 3.57ish for 3-4 hours a day. I do have a DIY 55ah li-ion 7s pack tied into the mix that wants to be at 4.2v, so maybe I'm not floating at all.

 

[Ampster]

Perhaps you do not have an LFP battery and have another Lithium chemistry which needs a higher voltage. The title said LifePO but if your battery is not LifePO then we should start over with a new thread or change the title and/or figure out what you actually have?

 

[Substrate]

Saw that! Was going to say there's NO wiggle room with non-lfp. Appears that you are "tieing it in" somehow? Danger Will-Robinson.

If you are reaching full charge by noon and floating, *with LFP*, then you have the luxury of dropping your CV voltage a bit - no need to reach full charge so early.

But yeah, if this is really about Li-Ion, well it's a whole different story.

 

[Ampster]

But yeah, if this is really about Li-Ion, well it's a whole different story.

That is the danger of jumping into the middle of a thread with out introducing oneself or making sure one is on the right thread.

 

[HighDesertOffgrid]

Saw that! Was going to say there's NO wiggle room with non-lfp. Appears that you are "tieing it in" somehow? Danger Will-Robinson.

700 ah of lifepo4 and 55 ah of li-ion all in the same system. Guess I live on the edge. lol

 

[HighDesertOffgrid]

That is the danger of jumping into the middle of a thread with out introducing oneself or making sure one is one the right thread.

With the exception of your misunderstanding of a brief mention of the li-ion, all of my comments have been on topic. I am sure I'm on the right thread.

 

[Substrate]

Yeah, sorry about that - my replies get shorter as the coffee wears off.

We'll assume you know the difference, and are very careful, using different systems for the different chemistries, but tying them together in the same charging system would just be problematic in my mind.

Looks like it's pretty well covered by the others earlier.

 

[HighDesertOffgrid]

Yeah, sorry about that - my replies get shorter as the coffee wears off.

We'll assume you know the difference, and are very careful, using different systems for the different chemistries, but tying them together in the same charging system would just be problematic in my mind.

Looks like it's pretty well covered by the others earlier.

I didn't intend to hijack the thread with my frankensystem. I have been mixing them for about 18 months without issue. But I also ride a motorcycle 90mph between lanes of rush hour traffic, so there's that.
Back to the question at hand, the paper linked above states "long-term" floating is the problem, not floating for a few hours daily.