LFP's Charging Voltage and State of Charge

[john61ct]

Yes, but the capacities you cite are I think a bit low.

Everything depends on how the owner **defines** 100% Full.

Take a given "360AH rated" bank, trusted supplier, top QA run from CALB.

That 360AH could be called "rating 100%" SoC if you like, benchmark A

Say you top-balance at 3.65V

Use that setpoint and hold Absorb until acceptance completely stops, let rest 24hrs, call that definition "vendor 100%" or benchmark B.

Run a standard load test to 12V, and you may well find B's capacity is 380AH, compared to A's 360AH.

Even at a 125A charge rate (.3C), getting to a 3.45Vpc setpoint and Just Stop - **my** definition of 100% SoC, call it benchmark C - may well load-test out at 345AH.

Now, as you say, the fact that that is 91% of the B definition is not really relevant to me, since that benchmark is based on a charge profile I consider harmful to longevity, not to be considered for normal daily cycling.

Same with the fact that "my C definition" of 100% is 96% of the A rating.

Now, lets get away from Just Stop no Absorb charging, keeping my 3.45V setpoint, but holding Absorb until acceptance completely stops. This according to Maine Sail's testing here http://www.cruisersforum.com/forums/....php?p=2731306 will get you to over 99.9% of benchmark B and thus 106% of rated A 360AH.

Same voltage same current, stopping at an endAmps setpoint of .06C 22A, gets you to say 99.1% of B, 104% of A.

But Cpt Pat's point is that instead of a 3C rate, if your only charge source is a few hundred watts of panels, your rate may be under 10A going into the bank, or less than .03C, and that rate is constantly varying with changing conditions.

If you are wanting to Hold Absorb at your voltage setpoint, as Cpt Pat does, you no longer have a reliable endAmps measure.

But dropping the setpoint to Cpt Pat' 3.40V changes things dramatically. I have not load tested any benchmarks there, but MaineSail's tests show even holding Absorb until acceptance stops completely, won't get SoC above 90% compared to benchmark B, which is 95% of rating A.

Cpt Pat's (and I believe MS may still agree?) major point is that even at that "low" voltage, once the bank is "Full" - however that is defined - continuing to push current into the bank for extended periods past that is "overcharging", in other words harmful, reducing its longevity.

However, many owners with solar-only are struggling to get to **any** definition of Full, so this would rarely be of concern for them.

Plus, the sun always sets.

Therefore, unless the bank was Full at noon and no loads are drawing on the bank, the Absorb Hold Time is not going to be very extended.

So, without being dogmatic about it, so far I tend to agree that relying on an SoC meter to end the charge cycle may not be needed in many use cases, even at low charging rates.

Constructive feedback welcomed. . .

[CatNewBee]

Well there are lots of theories about SOC, voltages and concerns about life expectancy - and on the other side successfull implementations and flawless running systems.

I admit, like any other technical system, anything can fail.

But it is like in the movie "War games" - the only safe bet to win is not to play - that is what most people with concerns about LFP do, very few have actually practical experience with the cells, they just look for excuses not to buy a LFP battery and try it themselves, instead scaremongering the others...

But there are also lots of early adopter and DIY installations, that just work everyday with no issues and no manual intervention / manual monitoring / manual balancing. LFP cells are quite robust anyway.

Finally it is your decision if you keep speculating or start doing it. You will learn a lot on the way and become finally more relaxed. Just be a little conservative and stick to the specs.

[Bob James]

So two things to mention .
It is preferable to charge LiFePo4 and in fact any Lithium based battery slowly. If you charge them at ridiculously high values like 3C you will plate the plates with lithium metal.

Second here is an example of one of my systems running perfectly for 12 months and I have another that is the same and has been running for two years now flawlessly .My charging stops at 27.4V . Top voltage on 8 cells is 28.4V and is never exceeded after the initial top balance of cells to 3.7V. and equalisation overnight.

[john61ct]

Yes, even just 1C is running hot, safety requires crazy infrastructure for a big bank, and reduces longevity.

As stated my usual is .3C, .5C now as high as I'll go even if in a real hurry.

[CatNewBee]

Even .3C or .5C are ways higher than charging currents in FLA technology. On a decent house bank you rarely would have the charging capacity of 1C.

I have 1000Ah, it would be possible to charge with 1000A, but no source delivers 15kW as a charger nor does the shore cable support that load (130A at 120V or 65A at 240V).

So this is a theoretical limitation, except you deal with small cells below 100Ah and realy big charger.

[Cpt Pat]

Originally Posted by john61ct

Yes, even just 1C is running hot, safety requires crazy infrastructure for a big bank, and reduces longevity.

As stated my usual is .3C, .5C now as high as I'll go even if in a real hurry.

Good advice. Heat is the primary enemy of LFP longevity. Storing at a high SoC is a distant second. The heat generated during charging/discharging increases by the square of the current (P = I squared R). Double the current, quadruple the fraction of power dissipated as heat.


We should avoid the abuses that occur in electric vehicle (EV) applications. The LFP manufacturers like to tout how fast their batteries can be charged when catering to EV users, but they don't stress the negative impact that practice has on battery life.


I have the same limits: 0.3C charge and discharge. Charging at 0.3C generates only 9% the amount of heat as charging at 1.0C. (0.3 squared X 100).

[john61ct]

On land I want to reverse-engineer the vehicle charge station output or figure out the signaling.

But only for stations where I pay for the energy I take of course, as in Austin.

Also the hybrid EVs make really nice gensets, can output a fair current even stepped down to 12V nominal.

But I digress 8-)

[Cpt Pat]

Originally Posted by john61ct

Yes, but the capacities you cite are I think a bit low.

Everything depends on how the owner **defines** 100% Full.

Take a given "360AH rated" bank, trusted supplier, top QA run from CALB.

That 360AH could be called "rating 100%" SoC if you like, benchmark A

Say you top-balance at 3.65V

Use that setpoint and hold Absorb until acceptance completely stops, let rest 24hrs, call that definition "vendor 100%" or benchmark B.

Run a standard load test to 12V, and you may well find B's capacity is 380AH, compared to A's 360AH.

Even at a 125A charge rate (.3C), getting to a 3.45Vpc setpoint and Just Stop - **my** definition of 100% SoC, call it benchmark C - may well load-test out at 345AH.

Now, as you say, the fact that that is 91% of the B definition is not really relevant to me, since that benchmark is based on a charge profile I consider harmful to longevity, not to be considered for normal daily cycling.

Same with the fact that "my C definition" of 100% is 96% of the A rating.

Now, lets get away from Just Stop no Absorb charging, keeping my 3.45V setpoint, but holding Absorb until acceptance completely stops. This according to Maine Sail's testing here "Floating" LFP batteries - Page 6 - Cruisers & Sailing Forums will get you to over 99.9% of benchmark B and thus 106% of rated A 360AH.

Same voltage same current, stopping at an endAmps setpoint of .06C 22A, gets you to say 99.1% of B, 104% of A.

But Cpt Pat's point is that instead of a 3C rate, if your only charge source is a few hundred watts of panels, your rate may be under 10A going into the bank, or less than .03C, and that rate is constantly varying with changing conditions.

If you are wanting to Hold Absorb at your voltage setpoint, as Cpt Pat does, you no longer have a reliable endAmps measure.

But dropping the setpoint to Cpt Pat' 3.40V changes things dramatically. I have not load tested any benchmarks there, but MaineSail's tests show even holding Absorb until acceptance stops completely, won't get SoC above 90% compared to benchmark B, which is 95% of rating A.

Cpt Pat's (and I believe MS may still agree?) major point is that even at that "low" voltage, once the bank is "Full" - however that is defined - continuing to push current into the bank for extended periods past that is "overcharging", in other words harmful, reducing its longevity.

However, many owners with solar-only are struggling to get to **any** definition of Full, so this would rarely be of concern for them.

Plus, the sun always sets.

Therefore, unless the bank was Full at noon and no loads are drawing on the bank, the Absorb Hold Time is not going to be very extended.

So, without being dogmatic about it, so far I tend to agree that relying on an SoC meter to end the charge cycle may not be needed in many use cases, even at low charging rates.

Constructive feedback welcomed. . .

I agree with the substance of your post. And I use your method of defining: "vendor 100%" or benchmark B. That's where I calibrate "100% SoC" on my totalizer. I recalibrate every few months, and thus far I've never seen a calibration/capacity drift more than 5% -- after tweaking the charge efficiency (98%) and Peukert function (1.02) settings.

Explaining what I meant:

Cpt Pat's (and I believe MS may still agree?) major point is that even at that "low" voltage, once the bank is "Full" - however that is defined - continuing to push current into the bank for extended periods past that is "overcharging", in other words harmful, reducing its longevity.

I mean charging once the battery reaches the actual maximum possible 100% SoC. What you call "'vendor 100%' or benchmark B." Not a user-defined maximum defined below that value, what you call "benchmark C."

Cpt Pat's (and I believe MS may still agree?) major point is that even at that "low" voltage, once the bank is "Full" - however that is defined - continuing to push current into the bank for extended periods past that is "overcharging", in other words harmful, reducing its longevity.

Yes. Well stated. That is my major point.

However, many owners with solar-only are struggling to get to **any** definition of Full, so this would rarely be of concern for them.

Plus, the sun always sets.

True for the specific case you cite, but... I have a combination of solar and a towed-impeller hydrogenerator (TIH). If my vessel is at or near hull speed, that TIH generates 8 amps continually -- 24 hours a day, while my consumption is 4 to 5 amps (an amp higher at night). Others have wind turbines that can deliver the same type of continual charge. Without monitoring SoC, I could easily overcharge my LFE battery with that 0.05C charging current. And since seastate causes the output of the TIH to continually fluctuate, I can't detect a charge current ramp-down.

By the way, with only my lead-acid batteries, it's disappointing to see the current output from my TIH drop at around a terminal voltage of 13.5 volts. But with my LFP on line, it's a thing of beauty to not see that current taper off. It happily soakes up whatever I can give it - I get full use of the TIH all the way up to charge cutoff! Maybe I'm easily entertained, but I have been very happily engaged just watching the ammeter!

In practice, my solar panels allow me to be a power hog during the day, running my stereo full blast (200 watts!) to annoy any nearby submarines I used to be a submariner - and yes - it can be annoying to whoever is listening to passive sonar from 50 miles away. So far, no one's surfaced to tell me to turn it down (or wasted a torpedo on my tiny boat).

[Delfin]

Originally Posted by Bob James

I'd have to take you to task on that.
When Landrovers came out in the war and just before they were all positive earth ,as were most cars . Shortly thereafter most changed to negative earth . Reason? Frankly the body stays charged up and that assists rust . Believe it or not is up to you but I know its true because I still have one with a damn good chassis and steel work under the alloy sheet skin. Now thats 70 odd years old and runs round the farm like it was a youngster. Outdoes the 4 WD tractors because it has little weight. You can today go buy a washing machine and find the back panel is made of masonite . They even paint it sometimes.

I'd also agree that avoiding the knees and you will be pleased is valid. But my concern is the heat that is produced in the center of the cell. That's what swells LFP because the center heat is the last to dissipate and it boils the small amount of electrolyte that is in the separators and you cant put it back in.

I have not seen any technical information from anywhere that would support your 40 % SOC claim.

I have two years only experience specifically in LFP and 24 V only at this point but my first LFP pack is as good as it was at the start, no corrosion ever and excellent voltage profiles. I have 5 battery pack installations on the go now.

The bms's are everywhere if you bother to look. I'v had them at 50 amps and not even warm . You are wrong. They make good stuff these days if you are careful and get one to test first

https://www.aliexpress.com/wholesale...ort+with+balan

So, if you disconnect charging at 28.3 volts, when that voltage is experienced by the battery for three minutes or, you think this is harmful? I don't think so....

And I have a BMS, likely one of the best, at least according to Mainesail, and I use it as the safety device it is designed to be.

Regarding storage at 40% SoC....https://batteryuniversity.com/learn/article/how_to_store_batteries

[newhaul]

Originally Posted by CatNewBee

Even .3C or .5C are ways higher than charging currents in FLA technology. On a decent house bank you rarely would have the charging capacity of 1C.

I have 1000Ah, it would be possible to charge with 1000A, but no source delivers 15kW as a charger nor does the shore cable support that load (130A at 120V or 65A at 240V).

So this is a theoretical limitation, except you deal with small cells below 100Ah and realy big charger.

so I will be just fine as long as my solar and wind are not putting out max at the same time
100ah Lfp with 400 watts solar and 400 wind. ( 60 amps potential) a .6c rate. We all know it will never happen.

[rgleason]

The Trojan Trilliam website now has a graph for #cycles base on C/2 or C and DoD. The number of cycles is higher of course for C/2, but it is unclear if this is for charge or discharge. Perhaps someone knows.

https://www.trojanbattery.com/trillium/


Also Trojan has hired Voltaig for testing of the new batteries
https://www.prnewswire.com/news-rele...c=eml_cleartim


https://www.voltaiq.com/

[john61ct]

Let's keep this thread about defining Full SoC, charging voltages, varying those for different current levels, and other issues raised by Cpt Pat's insights, especially as quoted in the OP

The impact of current on longevity is not in question. And LFP as a generic chemistry, not specific brands.

Otherwise these threads get so easily derailed, and much less useful.

Thanks

[rgleason]

Sorry but the title says "LFP's Charging .."


And actually I had not seen anything that quantifies the loss or gain of cycles dependent on C/2 or C so I thought and still think it is germane.


The Trojan reference was just to show this graph, and not intended to be a diversion.


So sorry if it "derails" a discussion, so I guess you are uninterested in the affect of rate of charging on longevity or cycles, however I am realizing it puts some additional practical limits on the choice of an alternator for LiFePo4.

Originally Posted by john61ct

Let's keep this thread about defining Full SoC, charging voltages, varying those for different current levels, and other issues raised by Cpt Pat's insights, especially as quoted in the OP

The impact of current on longevity is not in question. And LFP as a generic chemistry, not specific brands.

Otherwise these threads get so easily derailed, and much less useful.

Thanks

[Cpt Pat]

Originally Posted by rgleason

So sorry if it "derails" a discussion, so I guess you are uninterested in the affect of rate of charging on longevity or cycles, however I am realizing it puts some additional practical limits on the choice of an alternator for LiFePo4.

With apologies, I'm going to go down this slightly off topic tangent because I believe it brings up an interesting topic. Heat is the number one enemy of longevity, and the more cycles the more the cumulative degradation from heat.

Charging a LFP with an alternator that outputs relatively high charge rates may (and I suspect will) have a negative impact on LFP life. And here's the on topic part: Especially if you charge the cells to or above the 100% SoC point, where the partial energy converted to heat rises rapidly.

If the manufacturer rates the battery at a 2C charge rate, for example, I believe (though I haven't seen a manufacturer state this) that the battery could tolerate a higher rate electro-chemically, but at 2C and 100% SoC it is approaching a critical temperature rise over ambient (usually assumed to be 25 degrees C) where chemical or physical damage will begin to increasse precipitously. I base that on what I've studied about the LFP chemistry. At an elevated ambient temperature, the damage would occur at a lower charge rate, and continuing that (albeit error-prone inductive) logic, cooling the battery by blowing a fan on it would permit a higher rate - at least so far as the thermal factor is concerned. From a thermal engineering standpoint, stacking the batteries tightly together, as is generally done, worsens the thermal transfer characteristics (case to ambient) because it reduces the effective dissipative surface area and increases the thermal impedance. The cells in the middle of the stack would suffer the most.

At very low fractional currents, the heat rise would be almost irrelevant, but at high currents, it could become a life-limiting factor.

The heat generated increases by the square of the current, so a reduction of just half the current reduces the heat by a quarter. Reducing the current to 71% cuts the heat in half.

Unless truly abused, LFP batteries last so long in fractional-rate storage implementations that we don't have a good experience base for lifetime comparison, but I believe we can safely say that reducing the charge/discharge rate will pay some as-yet undefined positive benefit on longevity and cycle life. And in high C charging, blowing a fan on the stack (which consumes little current) won't hurt and will probably help. I cringe when I see the "drop in" LFP battery stacks that are enclosed in plastic cases. Those are nearly as resistive to heat transfer as thermos bottles.

It pays to be conservative - though how much -- I cannot say.

[CatNewBee]

Originally Posted by Cpt Pat

With apologies, I'm going to go down this slightly off topic tangent because I believe it brings up an interesting topic. Heat is the number one enemy of longevity, and the more cycles the more the cumulative degradation from heat.

... From a thermal engineering standpoint, stacking the batteries tightly together, as is generally done, worsens the thermal transfer characteristics (case to ambient) because it reduces the effective dissipative surface area and increases the thermal impedance. The cells in the middle of the stack would suffer the most....

Unless truly abused, LFP batteries last so long in fractional-rate storage implementations that we don't have a good experience base for lifetime comparison, but I believe we can safely say that reducing the charge/discharge rate will pay some as-yet undefined positive benefit on longevity and cycle life. ...

LFP batteries indeed suffer from heat, but they stay cool on charging and usage much more than FLA batteries, it is the ambient temperature that kills them. I know this, because my BMS has cell sensors and the battery stays at ambient temp almost regardless of load and charge. Just keep heat sources like engines, charger or inverter away.

Regarding the stacking, it is necessary and advised to stack the cells tight with some compression to prevent inflating the cells on deep discharges when they try to run off. This prevents faster irreversible mechanical damages to some degree, it is a safety issue. Also it enforces the blow of the safety vent and prevents a mess from blown up enclosures.

Also important to use the cells in the proper orientation to allow the fluid inside cover the pack properly, see manufacturer data sheet, do not lay them down horizontally, best is upright position of the battery poles.