"Floating" LFP batteries

[LoudMusic]

Originally Posted by 44'cruisingcat

13.8 trips absorb, timed for 1 minute, then float at 13.2.

Outback flexmax 80.

And is there a bulk phase?

[Delfin]

Originally Posted by tanglewood

I think this is at the heart of the debate. Our LA mind-set equates float with an on-going charge current, because that's what happens with LA. On-going charge current is harmful to LFP, therefore float must be bad for LFP, right?


But there is only a charge current if the float voltage is greater than the battery's resting voltage. In our past LA world, this is always the case, but in our new LFP world lowering the float voltage to the resting voltage of the battery results in ZERO charge current. It's the same as being disconnected. But the charger or alternator is floating along ready to carry any DC loads, and there is no need to switch out the LFP, and switch in some other battery.


Now I'm not saying there is anything wrong with other battery management approaches. You have a setup that works for you, and that's what matters. But I do think there is more to the whole float question, and it's not all bad. In fact, I think there are some attractive aspects to it if done correctly. I'm hoping to get all that out on the table, and better understood.

I do see your point. Where I come down on this is that one should set the automatic regulation such that if you slip into coma and can't disconnect the LFP bank from the charging sources, the settings won't damage the batteries. This presumes you have two banks, one of which is lead so charge sources have a destination. That said, I can't see how you could be wrong that if the float voltage is 13 or 13.1 v or thereabouts, the batteries could be stressed. But, my knowledge is limited to what I have read, and what I observe from my own LFP bank, so I defer to the experts. What I do know is that managing these banks is no where near as complicated as some make it out to be, assuming you can escape the LA mindset of expectations and just start treating the LFP bank as a fuel tank - fill, turn off the hose, use, re-fill.

[Delfin]

Originally Posted by LoudMusic

And is there a bulk phase?

Absorption voltage is the bulk voltage. The regulator maintains that bulk voltage setting for an "absorption" phase for a time period that hopefully can be adjusted, as Cat's Outback allows.

[john61ct]

I have not read the thread.

Floating at a low enough voltage will not murder an LFP bank.

However it will reduce longevity.

Yes even at such low voltage that there is no charge current flowing.

If an energy source is able to carry ongoing loads, the LFP bank should ideally be isolated from the circuit.

LFP is harmed by sitting at or even near Full. They should be kept at low SoC when not actively being cycled.

Ideally they should only be brought up from a low SoC to "Full" as in 3.45Vpc, **just before** a load will start drawing them down again.

Now practicality in some contexts may mean compromises from the above ideals.

But that will have a cost in lifetime cycles off the back end.

Say a bank is rated for 3000 cycles, but by being well coddled, avoiding the voltage shoulders you may ideally get 8000.

Maybe by Floating at 3.3Vpc for avg say 3 hrs per day you only get 7000.

All these numbers WAGs pulled out of my butt, but give an example of scale for those who want / need to Float, it is not a critical issue, but certainly an issue for those seeking to pass their LFP bank on to the grandkids.

8-)

[Nicholson58]

Do not float. This will damage the battery.

Read this. https://marinehowto.com/lifepo4-batteries-on-boats/

Chargers need to be re programmed to stop when any cell in the Battery assembly reaches full charge. In the event that they do not, the BMS must be able to disconnect the chargers. This can be a problem for some charge sources. My solar MPPT can only be remotely shut down by disconnecting all solar panels. My Balmar alternator controls can be programmed to shut down with new software. My generator must always be connected to a battery so it must have a fuel interrupt relay. My victron can be re-programmed if I download and install (successfully) new software. For all of these reasons, we will install Firefly carbon foam AGM replacements this season.

[Delfin]

Originally Posted by john61ct

All these numbers WAGs pulled out of my butt

Thanks John. I understand your methodolgy and basis for some of your recommendations better now.

[john61ct]

Fact is no one has more accurate numbers, and there are so many variables IRL they are not knowable.

Just trying to help, give a "big picture" perspective, don't want people thinking that a 13.2V Float will cut cycle lifetime in half.

Nor that just this one factor will magically extend lifetime while ignoring the others.

Staying under 3.5Vpc while charging in particular has a **much** more significant impact.

As does shallower discharge, and most people will still be happy "violating" that.

[44'cruisingcat]

Well the object of float is shallower discharge. Our bulk phase usually finishes around lunchtime. If I then disconnected the solar, all the loads for the rest of the day and night would come out of the battery.

By using float, most of the loads up till around sunset are carried by the solar.

I'll usually run the watermaker and hot water system with the solar in float.

[CraigOverend]

Conventional graphite negatives operate at a voltage only about 150 mV higher than that of lithium metal. The Solid-Electrolyte Interface (SEI) on those negatives has a certain resistance to passing lithium ions and if the charge current exceeds a value that would result in a 150 mV drop across the SEI, lithium ions will be deposited on the surface of the SEI instead of passing through. When not taken to extremes, this lithium plating process does not constitute a safety risk but nevertheless damages the cell and reduces its life.

The aging process in graphite negative electrodes involves a gradual ‘leakage’ of lithium ions through the SEI. The positive materials tend to release metallic ions (Mn, Fe, Ni and Co), which are reduced to metal on the negative, forming metallic clusters in the SEI that provide a conduit for lithium ions to pass through. These ions react with the solvents in the electrolyte and the resulting material is deposited on the surface of the SEI, which gradually thickens and increases its impedance. Furthermore, the lost lithium ions constitute the ‘fuel’ for the cell reaction so there is also a loss of capacity.

https://www.master-instruments.com.a...y___safts_.pdf


If I understand correctly, this seems to indicate 150 mV above open-circuit stability voltage may be the max you would want to float to with a graphite cathode cell.
I'd also only charge to the point that the oxygen evolution reaction (OER) begins to occur in the High-SOC section below, which will vary depending on cell chemistry. The same for oxygen reduction reaction (ORR) in the Low-SOC section below when discharging. Ideally you want to keep the cells in the Goldilocks zone where impedance is at a minimum, shown below as the Middle-SOC section.

If leaving the cells for an extended period, you probably want to set the float to the middle of the Middle-SOC voltage where there is the least impedance. Perhaps ultimately if you want your cells to last, you data log and monitor these and other points and have a battery management system that that adapts as the cells age.








Here's an actual LiFePO4 cell curve, note how abrupt the transition to OER at the end is. I'd charge that to 95% SoC and discharge to 15% and be using 80% capacity, whereas the one above I'd only go to 80% down to 20% and only be using 60% capacity.






And a ReLion 12V battery charge curve example can be found here. 10% to 90% depending on C-rate looks good for it.

[Maine Sail]

Originally Posted by tanglewood

The CALB recommended float voltage is 3.3 vpc. That's right at about 80% SOC, so that would be the equivalent storage SOC.

On the test bench currently (alpha testing for a battery monitor manufacturer) the used LFP bank, at 87.7% SOH was drawn down to a 12.50V cut off at a .1C discharge rate. Discharging to 12.50V, from 100% SOC, equated to final capacity of 8% SOC. The test ended in the middle of the night, I left auto charge off, and right now the OCV resting voltage has rebounded to 12.82V. So which voltage do we use to determine the safe standby/storage voltage during times at a dock?

I tend to define float for LFP as any voltage that exceeds the resting voltage after fully charging to 100% SOC. At voltages lower than natural resting 100% SOC I tend to call these storage voltages for no other reason than a term does not exist in the LA world and so as not to confuse a "standby voltage" with a LA type of floating..

If you charged to 100% SOC then allowed the bank to "discharge down to" a storage voltage of 3.3VPC. The graph below will give an idea of where you end up, even under load, after being fully charged then dropping to float.

Original Size Here: http://www.pbase.com/mainecruising/i...9/original.jpg


Original Size Here: http://www.pbase.com/mainecruising/i...0/original.jpg


We should be careful in interpreting LFP manuals so as not to confuse what they are trying to convey. The below image, if thinking in an LA charging mindset, would seem to mean that you can charge to 100% SOC then "drop to" 3.4VPC. That is what we do with LA to hold them at 100% SOC or darn close. As can be seen in the above two graphs dropping to 3.4V, after a full charge, would have you at about as close to 100% SOC as it gets.


When we read what they really meant it is charge to 3.4VPC, for certain standby applications, and we can then hold them at 3.4VPC. There is a difference between charging to 3.4VPC vs. charging to 100% SOC then dropping to 3.4VPC.



If you are using these cells in a standby system then they want you to charge to 3.4VPC / 13.6V not to 3.6VPC then drop back to 3.4VPC. If CALB has since lowered this to "charge to 3.3VPC" then this would mean charging only to 13.2V.

[Delfin]

Originally Posted by CraigOverend

https://www.master-instruments.com.a...y___safts_.pdf


If I understand correctly, this seems to indicate 150 mV above open-circuit stability voltage may be the max you would want to float to with a graphite cathode cell.
I'd also only charge to the point that the oxygen evolution reaction (OER) begins to occur in the High-SOC section below, which will vary depending on cell chemistry. The same for oxygen reduction reaction (ORR) in the Low-SOC section below when discharging. Ideally you want to keep the cells in the Goldilocks zone where impedance is at a minimum, shown below as the Middle-SOC section.

If leaving the cells for an extended period, you probably want to set the float to the middle of the Middle-SOC voltage where there is the least impedance. Perhaps ultimately if you want your cells to last, you data log and monitor these and other points and have a battery management system that that adapts as the cells age.








Here's an actual LiFePO4 cell curve, note how abrupt the transition to OER at the end is. I'd charge that to 95% SoC and discharge to 15% and be using 80% capacity, whereas the one above I'd only go to 80% down to 20% and only be using 60% capacity.






And a ReLion 12V battery charge curve example can be found here. 10% to 90% depending on C-rate looks good for it.

Very interesting, thank you...

[tanglewood]

Originally Posted by 44'cruisingcat

Well the object of float is shallower discharge. Our bulk phase usually finishes around lunchtime. If I then disconnected the solar, all the loads for the rest of the day and night would come out of the battery.

By using float, most of the loads up till around sunset are carried by the solar.

I'll usually run the watermaker and hot water system with the solar in float.

Bingo!! This is the goal. I want to cycle the LFP when there is no other source of power, not gratuitously when there is an alternate sort of power.

[tanglewood]

Originally Posted by Maine Sail

On the test bench currently (alpha testing for a battery monitor manufacturer) the used LFP bank, at 87.7% SOH was drawn down to a 12.50V cut off at a .1C discharge rate. Discharging to 12.50V, from 100% SOC, equated to final capacity of 8% SOC. The test ended in the middle of the night, I left auto charge off, and right now the OCV resting voltage has rebounded to 12.82V. So which voltage do we use to determine the safe standby/storage voltage during times at a dock?

I tend to define float for LFP as any voltage that exceeds the resting voltage after fully charging to 100% SOC. At voltages lower than natural resting 100% SOC I tend to call these storage voltages for no other reason than a term does not exist in the LA world and so as not to confuse a "standby voltage" with a LA type of floating..

If you charged to 100% SOC then allowed the bank to "discharge down to" a storage voltage of 3.3VPC. The graph below will give an idea of where you end up, even under load, after being fully charged then dropping to float.

Original Size Here: http://www.pbase.com/mainecruising/i...9/original.jpg


Original Size Here: http://www.pbase.com/mainecruising/i...0/original.jpg


We should be careful in interpreting LFP manuals so as not to confuse what they are trying to convey. The below image, if thinking in an LA charging mindset, would seem to mean that you can charge to 100% SOC then "drop to" 3.4VPC. That is what we do with LA to hold them at 100% SOC or darn close. As can be seen in the above two graphs dropping to 3.4V, after a full charge, would have you at about as close to 100% SOC as it gets.


When we read what they really meant it is charge to 3.4VPC, for certain standby applications, and we can then hold them at 3.4VPC. There is a difference between charging to 3.4VPC vs. charging to 100% SOC then dropping to 3.4VPC.



If you are using these cells in a standby system then they want you to charge to 3.4VPC / 13.6V not to 3.6VPC then drop back to 3.4VPC. If CALB has since lowered this to "charge to 3.3VPC" then this would mean charging only to 13.2V.

Thanks. It does appear to hinge on what you mean by "float". I like your term "standby voltage" to distinguish between a voltage that continues to create current flow into the batteries from a voltage that does not.


My concern is that "don't float" has become LFP dogma with people not understanding what's behind it. So nobody (or very few people) operates with chargers at a standby voltage to carry loads rather than cycling the batteries. This standby voltage is accomplished by carefully programming the "float" voltage and trigger conditions in a charger. But the use of that nasty word, "float" scares people away.


The broken english in the CALB manual is pretty priceless, and often very difficult to figure out what they are saying. I took that quoted section to mean the float voltage in applications where there is an on-going power source, unlike an EV or portable device. So precisely what we are now calling a standby voltage.


But you are saying it's a different charge protocol? And I'm not sure what you mean by "If you are using these cells in a standby system then they want you to charge to 3.4VPC / 13.6V...". Is it charge to 3.4 vpc and hold there? And if so, what's the "/ 3.6 vpc" part mean?


I think another factor in this is how full you charge the cells in the first place. Charging to 3.6 vpc would be full charge, so holding them there with a standby voltage would essentially be storing them at full charge which is discouraged. Charging to 3.4 vpc and holding there would be charging to what, around 70% SOC and holding there? That seems OK, and would be consistent with the described standby power applications. My assumption is that on a boat, batteries are operated roughly 20%-80% SOC, so a charge to say 3.45 vpc, followed by a standby voltage of 3.30 vpc. That would charge to about 80%, then drain down to maybe 70% and hold at that point.


The alternate strategy of disconnecting the LFP bank and switching in a small LA buffer bank of course works, but I'm hoping to eliminate as much switching and switching logic as possible. Plus I'd like to not be cycling my batteries every day when on shore power while my boat sits for a month or two or three.


Oh, and there is another reason I'd like to keep the LFP bank connected. My inverters have a boost function where they will draw battery power to supplement shore or gen power for short term overloads. I think this is an excellent way to live on smaller shore power connections that you might otherwise need, and to run a generator that is sized more to the average power load rather than the peak power load. None of this may matter to a typical sailing vessel, but for other types of boats it's a big benefit.

[a64pilot]

As a side note, would terminating charge with some really low voltage like 12V effect the battery at all?
Wouldn’t drain the bank, would it?
I’m not talking plugged into a dock and a permanent float, but as a way to terminate charge during the day with Solar, assuming only an hour or so of float before Solar drops off line anyway.

[Maine Sail]

Originally Posted by tanglewood

But you are saying it's a different charge protocol? And I'm not sure what you mean by "If you are using these cells in a standby system then they want you to charge to 3.4VPC / 13.6V...".

Is it charge to 3.4 vpc and hold there? And if so, what's the "/ 3.6 vpc" part mean?

It means charge only to 3.4VPC or 13.6V (for a 12V bank computation) for standby systems not charge to 3.6VPC then drop back to 3.4VPC.

Originally Posted by tanglewood

My assumption is that on a boat, batteries are operated roughly 20%-80% SOC, so a charge to say 3.45 vpc, followed by a standby voltage of 3.30 vpc. That would charge to about 80%, then drain down to maybe 70% and hold at that point.

Take a closer look at those graphs I posted. That is my own bank of 2009 Winston cells. That graph was cycle test 1000 on 9 year old cells. It took many years of cycling to 80% DOD to get there, 1000 cycles is a LOT.

For that capacity test the cells were charged to 13.8V or 3.45VPC and current allowed to taper to 10A. The bank, at 9 years old and 1000 cycles still delivered well in ecess of the 400Ah rating at over 416Ah's.

The cells deliver the same capacity if I charge them to 3.6VPC. The idea that charging to 3.45VPC can't get your bank to 100% SOC is just not the case. If you charge to 13.8V/3.45VPC, then stop immediately, you will be slightly below 100% SOC but if you absorb even less than 30 minutes at 3.45VPC the bank will reach 100% SOC. Charging to 3.45VPC then dropping to 3.3VPC is not a strategy I would employ on my own cells as seen in the graphs because it leaves you at 99.9% SOC from a full charge.

I just programmed in a charge to 3.4VPC test. I will then discharge to 0% SOC and let you know the SOC attained at charge up to 3.4VPC. We already know what charge to full then drop to 3.3VPC is, it's in the graphs...