"Floating" LFP batteries

[Delfin]

Originally Posted by Maine Sail

Where we cruise, Maine & Maritimes, there are so few marinas it's a non issue. That said flipping off the LFP and then flipping on the lead bank takes but a few seconds. It's more work getting out the shore power cable & plugging it in. The last time we used a marina we never plugged in at all. With lead I would have jumped at the opportunity to plug in. With LFP I just poured another cocktail and sat back and relaxed.......

Ditto. I use the ACR switch to direct charge current to the starter bank.

Step 1: Flip toggle on the ACR from Auto Combine to Manual Combine.
Step 2: Turn off BMS
Step 3: Prepare adult beverage

When this proves too much effort, I'll be ready for adult diapers.

[john61ct]

Originally Posted by tanglewood

Does a week at 100% SOC reduce life, or does it make no difference if you then fully cycle the battery?

Absolutely.

Minutes count, much less hours.

Not saying we know exactly how much lifetime is reduced, too many variables.

Also, one super important note: people toss around SoC% levels as if they have an objective meaning.

They don't.

_______
At the bottom, maybe we can all agree on 12V resting as a common definition of Zero SoC, as we do with lead at 10.5V.

Since going much lower **in voltage** risks instant scrapification, as MS indicated above, it would be useless to assign SoC% values down there.

Now assuming this is a good quality 12V/400AH rated bank with a CEF of 1.05.

Take it from a 0% SoC state, and:

Case A. add 420AH to it at a 200A rate at 14.6V

Case B. add 420AH at a 40A rate at 14.6V

Case C. bring it up to 13.75V and Just Stop (no Absorb)

Case D. bring it up to 13.8V and Absorb until endAmps of 12A is reached

Case E. charge to 14.6V and Absorb as with D

Assume E is the cell maker's published spec for 100% Full, but you agree with me that this is "pushing into the shoulder" **so far** as to drastically reduce longevity,

perhaps so the owner only gets the rated lifetime, rather than its far higher true potential, even if the bank is coddled every other way.

So why call that SoC 100%, if you're never going to go there in normal daily operations?

I personally use Case D as **my definition** of 100% Full when precision is required, e.g. calibrating a SoC gauge.

Case C is my charge-stopping point for day to day cycling usage, and I intentionally vary it a bit in case the memory effect is real. It is usually above 98% SoC compared to D.

Some people here are under the impression that Case C & D are far lower SoC compared to E, even talking cray-cray low numbers like 80-90% SoC.

It is a difference far less than that.

And my 100% SoC at Case D is higher than A or B (in every case so far), "just how much exactly?" seems a bit of a silly question to me.

If you want to go to the trouble of finding out, post back here and let us know.

Now, back to my larger point.

Can you understand why I feel how silly statements like

"LFP should be cycled between 20% and 80%, and therefore you are losing 40% of capacity"

are?

Or attempts to correlate precise working voltage in the mid ranges to varying arbitrary definitions of SoC?

Given the above example, I just reset my coulomb counter to 100% at Case D. That **is** 100%.

It then counts down using a 400AH capacity setting so I can see **roughly** where my SoC is at.

Since my entertainment circuit alarms then LVD cuts out at around 12.5V, other non-essential big-amp load circuits at say 12.3V,

only essential safety/navigation systems are left going down to single-digit SoC readings, and their "user space LVD" cuts out at 12.1V.

The lowest failsafe "BMS" level never needs to isolate the bank completely at 11.9V, unless those higher-level protections fail.

I in fact have greater than rated AH capacity using **those** admittedly arbitrary definitions of 0-100% SoC, in my case prioritizing longevity.

And even if I didn't, I just don't care how much capacity I might be "wasting", compared to what? damaging voltages?

With a bank from a quality maker, it's bound to be an insignificant amount anyway.

[Delfin]

Originally Posted by tanglewood

Does a week at 100% SOC reduce life, or does it make no difference if you then fully cycle the battery? What about a month? Or a year?

See MaineSail's test of the loss of capacity for a battery stored for a year at 100% SoC. From memory, a loss of a bit under 12% over that time period. I have no clue if this is a linear function, and I suspect it is not, but assuming that it is, on a 100 amp bank, a day stored full is going to cost you 1/3 of an amp in lost capacity. The minutes John is concerned about may reduce the capacity by .000023 amp per minute.

The week you asked about would reduce your bank by 1/4 of an amp. To avoid this precipitous loss of capacity, folks who actually know what they are talking about recommend storing the batteries at 50% SoC, in the sweet spot of lowest impedance.

[Delfin]

Incidentally, Lithionics recommends storing their product at 100% SoC for periods less than 3 months, and at 50% SoC for periods greater than 3 months.

http://lithionicsbattery.com/wp-cont...cations-R0.pdf

[a64pilot]

Originally Posted by Delfin

Ditto. I use the ACR switch to direct charge current to the starter bank.



Step 1: Flip toggle on the ACR from Auto Combine to Manual Combine.

Step 2: Turn off BMS

Step 3: Prepare adult beverage



When this proves too much effort, I'll be ready for adult diapers.

So you charge your lead bank at the same voltage as your LFP?

[Delfin]

Originally Posted by a64pilot

So you charge your lead bank at the same voltage as your LFP?

Yes. The starter bank delivers current for an instant to start the engines. Underway, the bulk voltage setting of 28 v is more than enough to replace that. Minimum absorption, then float at 26.6, which supplies house loads underway through the inverter drawing off the LA bank. I installed today a Smart Gauge to check SoC of that starter bank and may adjust float up a bit. Since I disconnect charging of the Lithium bank when the acceptance rate drops to around 5%C, I could set float to what the LA bank needs, but just to be safe I keep it at a harmless rate for Li if I forget to throw a switch when the Li bank is full.

[john61ct]

Originally Posted by Delfin

a day stored full is going to cost you 1/3 of an amp in lost capacity. The minutes John is concerned about may reduce the capacity by .000023 amp per minute.

Talk about WAGing 8-)

And capacity is not measured in amps. . .

[john61ct]

Originally Posted by Delfin

Incidentally, Lithionics recommends storing their product at 100% SoC for periods less than 3 months, and at 50% SoC for periods greater than 3 months.

http://lithionicsbattery.com/wp-cont...cations-R0.pdf

Well that's clearly wrong.

I think the guideline for LFP in general needs to be, ignore all advice from vendors 8-)

[tanglewood]

Originally Posted by Delfin

See MaineSail's test of the loss of capacity for a battery stored for a year at 100% SoC. From memory, a loss of a bit under 12% over that time period. I have no clue if this is a linear function, and I suspect it is not, but assuming that it is, on a 100 amp bank, a day stored full is going to cost you 1/3 of an amp in lost capacity. The minutes John is concerned about may reduce the capacity by .000023 amp per minute.

The week you asked about would reduce your bank by 1/4 of an amp. To avoid this precipitous loss of capacity, folks who actually know what they are talking about recommend storing the batteries at 50% SoC, in the sweet spot of lowest impedance.

MainSail's test is an excellent data point, but without knowing the actual mechanism of capacity loss, I don't think we can conclude that it's a day for day linear loss for every day at 100% SOC. That may be exactly right. But at the same time maybe it's like LA where there is a chemical reaction that is completely reversible unless it's left sitting for too long? I just wonder if we really know, or if we are connecting dots that might not actually connect in a straight line. There may well be some scientific papers that have looked into this, but I haven't found them yet.

[tanglewood]

Originally Posted by john61ct

Absolutely.

Minutes count, much less hours.

Not saying we know exactly how much lifetime is reduced, too many variables.

Do we really know this? I would love to see the studies behind it. Have you found any? I'm not saying it's incorrect. I'm just trying to validate some of these beliefs/assertions.

[tanglewood]

Originally Posted by john61ct

_______
At the bottom, maybe we can all agree on 12V resting as a common definition of Zero SoC, as we do with lead at 10.5V.

Since going much lower **in voltage** risks instant scrapification, as MS indicated above, it would be useless to assign SoC% values down there.

Now assuming this is a good quality 12V/400AH rated bank with a CEF of 1.05.

Take it from a 0% SoC state, and:

Case A. add 420AH to it at a 200A rate at 14.6V

Case B. add 420AH at a 40A rate at 14.6V

Case C. bring it up to 13.75V and Just Stop (no Absorb)

Case D. bring it up to 13.8V and Absorb until endAmps of 12A is reached

Case E. charge to 14.6V and Absorb as with D

Assume E is the cell maker's published spec for 100% Full, but you agree with me that this is "pushing into the shoulder" **so far** as to drastically reduce longevity,

perhaps so the owner only gets the rated lifetime, rather than its far higher true potential, even if the bank is coddled every other way.

So why call that SoC 100%, if you're never going to go there in normal daily operations?

I personally use Case D as **my definition** of 100% Full when precision is required, e.g. calibrating a SoC gauge.

Case C is my charge-stopping point for day to day cycling usage, and I intentionally vary it a bit in case the memory effect is real. It is usually above 98% SoC compared to D.

Some people here are under the impression that Case C & D are far lower SoC compared to E, even talking cray-cray low numbers like 80-90% SoC.

It is a difference far less than that.

And my 100% SoC at Case D is higher than A or B (in every case so far), "just how much exactly?" seems a bit of a silly question to me.

If you want to go to the trouble of finding out, post back here and let us know.

Now, back to my larger point.

Can you understand why I feel how silly statements like

"LFP should be cycled between 20% and 80%, and therefore you are losing 40% of capacity"

are?

Or attempts to correlate precise working voltage in the mid ranges to varying arbitrary definitions of SoC?

Given the above example, I just reset my coulomb counter to 100% at Case D. That **is** 100%.

It then counts down using a 400AH capacity setting so I can see **roughly** where my SoC is at.

Since my entertainment circuit alarms then LVD cuts out at around 12.5V, other non-essential big-amp load circuits at say 12.3V,

only essential safety/navigation systems are left going down to single-digit SoC readings, and their "user space LVD" cuts out at 12.1V.

The lowest failsafe "BMS" level never needs to isolate the bank completely at 11.9V, unless those higher-level protections fail.

I in fact have greater than rated AH capacity using **those** admittedly arbitrary definitions of 0-100% SoC, in my case prioritizing longevity.

And even if I didn't, I just don't care how much capacity I might be "wasting", compared to what? damaging voltages?

With a bank from a quality maker, it's bound to be an insignificant amount anyway.

I get your SOC point.


I think most people start with the battery manufacturers definition of Capacity and SOC, and work from there. I expect manufacturer's capacities will be on the high side, charging and discharging to aggressive limits so they can claim higher capacities. But those numbers are also paired with an expected cycle life, so they don't want to be too aggressive or their advertised cycle life will suffer.


From the mfg defines capacity and SOC, each of us decides how far we want to discharge and charge with a goal of longer battery life, and lower risk of overshoot and damage. That yields a smaller "working capacity" and a new "working SOC" range, with new "working" discharge and charge IV points.


I agree with you that in practice, these new, personalized "working" parameters become the only ones that matter, and are inherently our "new" 0-100% SOC range.

[Delfin]

Originally Posted by tanglewood

MainSail's test is an excellent data point, but without knowing the actual mechanism of capacity loss, I don't think we can conclude that it's a day for day linear loss for every day at 100% SOC. That may be exactly right. But at the same time maybe it's like LA where there is a chemical reaction that is completely reversible unless it's left sitting for too long? I just wonder if we really know, or if we are connecting dots that might not actually connect in a straight line. There may well be some scientific papers that have looked into this, but I haven't found them yet.

I suspect it isn't, but don't know that it is not linear, as I noted. As I understand it, the chemical reaction is a loss of Li ions which reduces capacity, but I have no way of knowing the ramp of that process. One would assume it is not linear based on Lithionics data, who recommend storing at full charge, but only for 3 months, then at 50% thereafter with periodic cycling. Unless you believe, as some apparently do, that manufacturers think it is in their best interest to kill their products by recommending practices that will ruin their brand, I would assume Lithionics knows something more about the subject than I do. That said, it is easy to just not charge the Li bank on the day we return to the dock from the previous day's usage, so storing them at half SoC is no issue, so why not? Further, that is the recommendation of MaineSail who has probably spent more time at the test bench actually gaining empirical data on the batteries as well as using them personally than all of us put together. Seems a better source of information than from people whose sole source of information appears to be their alimentary canal.

You might give the Lithionics engineers a call. I found them very forthright and willing to tell you what they know and don't know and they might be able to point you in the direction of outside research on subjects of interest. After all, it's what they do for a living. I learned a lot from them.

[Delfin]

Originally Posted by john61ct

Talk about WAGing 8-)

And capacity is not measured in amps. . .

Thank you for the correction John. I should have written amp hour, not used shorthand to avoid confusion.

[magnificat]

The arguments seem to be irrelevant because lead acid batteries and lithium batteries are so poles apart in both expense and life. Safety requirements in Australia stipulate AGM for Cat 4 and above.. I have Varta (German) deep cycle house batteries that are in their 10th year still going strong. The boat is on a swing mooring, has no solar, no generator and uses shore power overnight each week after a days racing or after a weekend use. The batteries never lose charge at rest even for long periods.

[john61ct]

Well getting on shore power every night is very different from the use cases discussed here.

No challenge at all.