Float of LFP

[Captain Bill]

Originally Posted by s/v Jedi

The reason I set absorption at 13.8V (well, I don’t because I have a 24V system, but to keep it in your 12V system) is that the battery gets fully charged. The absorption time is tuned to get that full charge.

The reason for the float voltage at 13.15V is so that the battery discharges a couple percent before the MPPT take over. You want to see all current for loads come from the battery to bring it down a bit before letting the solar float take over.

What you write about not wanting to waste solar isn’t what happens with float at 13.15V and you will see that MPPT takes 100% of the load as soon as the battery has come down a couple percent.

This isn’t really a float stage, you hack the float stage to power loads from solar after the battery is already full but before the sun goes down. I find a float at 13.4V doesn’t take the pressure off the battery quickly enough.

To not waste solar energy after the battery is full, you need to consume more energy; keeping a high float on the battery doesn’t help at all because it’s already full. So start a water heater or watermaker, bake a bread etc.

I don’t have charge or re-engage charge parameters for my BMS. A BMS shouldn’t charge a battery imo.

Jedi, did you have a good reason for choosing 13.15 for your float (I assume really 26.3 in your case) or did you just pick that voltage? I use 13.3 because it's 90% SOC on every chart I have ever seen. 13.15 is closer to 50% SOC. On sunny days I have plenty of solar to carry most house loads in the afternoon and 13.3 keeps my reserve up in case of a cloudy day. Just wondering about your reasoning.

[s/v Jedi]

Originally Posted by Captain Bill

Jedi, did you have a good reason for choosing 13.15 for your float (I assume really 26.3 in your case) or did you just pick that voltage? I use 13.3 because it's 90% SOC on every chart I have ever seen. 13.15 is closer to 50% SOC. On sunny days I have plenty of solar to carry most house loads in the afternoon and 13.3 keeps my reserve up in case of a cloudy day. Just wondering about your reasoning.

You can’t use those SOC vs Voltage charts for this. When the battery is being charged, it is not in a state that allows comparing voltage with the chart. This is only allowed when the battery is disconnected from all circuits, both charge and discharge, for a “while”. Often they state hours, but after 10 minutes you get a pretty good idea.

This is even more difficult with LiFePO4 because it’s graph is very flat in the middle, i.e. there are very small voltage differences for very large SOC differences.

The only correct way to determine SOC during the charging process is by counting Ah going in/out the battery. This means you need a reference to start with, for which we use a fully charged battery and we call that 100% SOC so we reset the battery monitor to 100% and start discharging. When SOC is down to something like 30-60% you start the charge at my recommended voltages. Keep track of SOC and voltage and write that down for every 5% SOC. For the last 10% write it down for every 1%. Now you know at which SOC% your charger finishes.

You raise/lower absorption voltage aiming at a 99% SOC. If you charge slower than me, you may well need a lower voltage even.

This isn’t the only way to tune the charge process: I also work with the absorption time setting. If you end up at 100% you can try to lower it and when you end up at 98% increase it a bit. It’s like the fine tune function.

So don’t assume SOC%, you really need to measure it either with a Victron BMV or a good BMS. Note that the Victron BMV needs it’s “charged voltage” setting adjusted to just below the absorption voltage of the charger. This makes it reset to 100% when charging finishes, even though you finish just short of 100%. This is needed to recalibrate the BMV in order to keep accuracy. Check the manual for recommendations on this setting.

[Chotu]

^^^^. Nice post. Some solid advice and a a good technique to follow!

[barcoMeCasa]

Originally Posted by Captain Bill

Jedi, did you have a good reason for choosing 13.15 for your float (I assume really 26.3 in your case) or did you just pick that voltage? I use 13.3 because it's 90% SOC on every chart I have ever seen. 13.15 is closer to 50% SOC. On sunny days I have plenty of solar to carry most house loads in the afternoon and 13.3 keeps my reserve up in case of a cloudy day. Just wondering about your reasoning.

I went 13.3v from watching what was actually going on >

Cells would discharge a little when fridge ran then a little current would go back in. With bulk set to end @ 13.8v looks like soc would be pretty close to the top as C rate going in is very low, though this graph is for different cells.

Absorption more than a few minutes is pretty pointless, mA going in so I don't keep the voltage up there, nothing really to be gained.
Going lower with "float" seems pointless as well, can't find actual at rest fully charged voltage for my catl cells but all google can find shows at rest voltage of near full as being higher than 3.25v per cell. Seems pointless going any lower and the data backs this up, with no loads or charging the cells just sit there doing nothing.
Bear in mind these voltages are from a smartshunt close to the batteries & the victron regulators use that voltage for float etc. And this is just one boat... measure log whats really happening on yours

IMHO there are better resources online fior lifepo4 research than a sailing forum, nordkyn very well researched > https://nordkyndesign.com/charging-m...battery-banks/
lots in https://diysolarforum.com/ and offgrid garage on youtube has masses of real world resting.
Hope this at least might be interesting..

[s/v Jedi]

Originally Posted by barcoMeCasa

I went 13.3v from watching what was actually going on >

Cells would discharge a little when fridge ran then a little current would go back in. With bulk set to end @ 13.8v looks like soc would be pretty close to the top as C rate going in is very low, though this graph is for different cells.

Absorption more than a few minutes is pretty pointless, mA going in so I don't keep the voltage up there, nothing really to be gained.
Going lower with "float" seems pointless as well, can't find actual at rest fully charged voltage for my catl cells but all google can find shows at rest voltage of near full as being higher than 3.25v per cell. Seems pointless going any lower and the data backs this up, with no loads or charging the cells just sit there doing nothing.
Bear in mind these voltages are from a smartshunt close to the batteries & the victron regulators use that voltage for float etc. And this is just one boat... measure log whats really happening on yours

IMHO there are better resources online fior lifepo4 research than a sailing forum, nordkyn very well researched > https://nordkyndesign.com/charging-m...battery-banks/
lots in https://diysolarforum.com/ and offgrid garage on youtube has masses of real world resting.
Hope this at least might be interesting..

Unfortunately you don’t have a SOC% graph, just voltage and current. Are you missing this crucial data?

[barcoMeCasa]

Originally Posted by s/v Jedi

Unfortunately you don’t have a SOC% graph, just voltage and current. Are you missing this crucial data?

No, SOC gets recorded. Just don't see it as "crucial" cos it will be wrong straight away. From the masses of data and testing available on the web I'm happy that charging to the same place near enough up the top knee will mean the same soc every time, I never go 100%, seems pointless putting extra stress on them every day for no real benefit.
What this does mean though is having to manually set the smartshunt SOC now and then, even with efficiency set to 100% & peaukerts set to 1 it will loose about 0.6%/0.7% SOC each day even if the regulators hit the top of bulk @ 13.8v (smartshunt voltage) every day. I just use 85% for no real reason, seems a nice number. Likely low. After a while it gets a bit annoying so will get manually reset back to 85%.


imho it's a bit like boat navigation, really doesn't matter knowing exactly where you are every minute of the day, what is VITAL is knowing where you're not!.
What I see as crucial is keeping the voltage away from the ends of the knees, if soc is wrong but ensures that then fine.
It's been weeks now since mine were "full" going up the top knee, a 6 day goosewinged south passage meant sun behind the sails half the time & they've not crept back up yet, short days as well. They only ever get charged from solar. SOC couple of mornings has sneaked just under 30% but I know that's wrong, in the right way, it will be well above that. Could be 40%...60%, doesn't matter & will be a nice surprise one day seeing float again. What does matter is that it's safely well away from going near the bottom voltage knee which is the crucial bit.
Been fulltime living mostly on the hook for donkeys years now, it's so instinctively tempting after living with grumpy needy lead acid for so long to stick the honda on all afternoon to pull them back up but no, they're fine. Data don't lie.

Somehow suspect you might not agree... Your boat , your call .

[s/v Jedi]

Originally Posted by barcoMeCasa

No, SOC gets recorded. Just don't see it as "crucial" cos it will be wrong straight away. From the masses of data and testing available on the web I'm happy that charging to the same place near enough up the top knee will mean the same soc every time, I never go 100%, seems pointless putting extra stress on them every day for no real benefit.
What this does mean though is having to manually set the smartshunt SOC now and then, even with efficiency set to 100% & peaukerts set to 1 it will loose about 0.6%/0.7% SOC each day even if the regulators hit the top of bulk @ 13.8v (smartshunt voltage) every day. I just use 85% for no real reason, seems a nice number. Likely low. After a while it gets a bit annoying so will get manually reset back to 85%.


imho it's a bit like boat navigation, really doesn't matter knowing exactly where you are every minute of the day, what is VITAL is knowing where you're not!.
What I see as crucial is keeping the voltage away from the ends of the knees, if soc is wrong but ensures that then fine.
It's been weeks now since mine were "full" going up the top knee, a 6 day goosewinged south passage meant sun behind the sails half the time & they've not crept back up yet, short days as well. They only ever get charged from solar. SOC couple of mornings has sneaked just under 30% but I know that's wrong, in the right way, it will be well above that. Could be 40%...60%, doesn't matter & will be a nice surprise one day seeing float again. What does matter is that it's safely well away from going near the bottom voltage knee which is the crucial bit.
Been fulltime living mostly on the hook for donkeys years now, it's so instinctively tempting after living with grumpy needy lead acid for so long to stick the honda on all afternoon to pull them back up but no, they're fine. Data don't lie.

Somehow suspect you might not agree... Your boat , your call .

I fully get what you write, but you configured the SmartShunt incorrectly! Just a couple taps to get it perfect: in SmartShunt settings, find the “charged voltage” setting. Take your choosen absorption voltage and subtract 30mV and enter that value for “charged voltage” so if one has 13.8V as absorption voltage, then set this at 13.75V or so.

Now check that every day when the charger switches to float, SOC% is set to 100%. You know that in reality it is 99% or whatever you choose as your upper limit, but now the SOC% is fully accurate for the range you cycle in.

I expect that this will let you see my posts in a new light, why I recommend using relays programmed on SOC for storage mode and to use SOC for accurate charge termination.

I am at the same absorption voltage so we do the same but my SOC% is accurate.

[s/v Jedi]

Let me add how the BMV or SmartShunt uses this “charged voltage” setting:

During charge, current flows into the battery and the voltage will slowly increase until we hit the absorption voltage (I call it Constant Current, CC but most call it bulk phase).

At that moment voltage will be held steady at the set absorption voltage and current will still flow into the battery but slowly decreases. We call this the absorption phase (Constant Voltage, CV).

Now there is a timer running or current is monitored or both to determine when to stop this phase, at which time we go to float voltage, which is lower.

The Victron algorithm looks at voltage and current. When the voltage goes higher than the “charged voltage” setting, it pulls the hammer back, waiting for the trigger to be pulled. This trigger is the current: as soon as current flows out of the battery instead of into the battery, the algo fires and SOC is set to 100%.

This is why “charged voltage” must be lower than the absorption voltage. If it isn’t, the hammer isn’t pulled back and the trigger can’t fire.

[nfbr]

This video offgridgarage is why I view reliably chasing SOC charging as near impossible. You can only chase the knees, and the knee is 99%.
https://youtu.be/pijPu7t-akM?si=hn6BfjSQBvsVmRNd

Tesla found the same problem when they started with LiFePO4. Incorrect range (SOC) calibration due to no regular full charge. They never had the problem in 10 years of NMC, but had to change for LiFePO4z

Guidance now for NMC is 80% SOC charge regularly

LiFePO4 should be 100% at least weekly for range (SOC) calibration. Onlt way to find it is the knee. And thats at >99%

https://teslanorth.com/2022/03/09/te...batteries-pic/

[nfbr]

I do wish the 712 and the MPPT’s shared a common profile on that.
They dont even use the same presets

[barcoMeCasa]

Originally Posted by s/v Jedi

I fully get what you write, but you configured the SmartShunt incorrectly! Just a couple taps to get it perfect: in SmartShunt settings, find the “charged voltage” setting. Take your choosen absorption voltage and subtract 30mV and enter that value for “charged voltage” so if one has 13.8V as absorption voltage, then set this at 13.75V or so.

Now check that every day when the charger switches to float, SOC% is set to 100%.

Do remember having a good look at autoresetting but really didn't like it. 13.8v with no absorption will be below 100% soc & having the error go the "wrong way" just felt all wrong. Offgrid garage tested 3.4v end of bulk resulting in 89% SOC result. That was at 0.2c so solar with much less c & 3.4c pc will be higher but even so, much prefer to have any errors being on the safe side.. No big deal anyway, it's only a display, the usefulness comes from daily in/out.

[CaptainRivet]

Originally Posted by nfbr

This video offgridgarage is why I view reliably chasing SOC charging as near impossible. You can only chase the knees, and the knee is 99%.
https://youtu.be/pijPu7t-akM?si=hn6BfjSQBvsVmRNd

Tesla found the same problem when they started with LiFePO4. Incorrect range (SOC) calibration due to no regular full charge. They never had the problem in 10 years of NMC, but had to change for LiFePO4z

Guidance now for NMC is 80% SOC charge regularly

LiFePO4 should be 100% at least weekly for range (SOC) calibration. Onlt way to find it is the knee. And thats at >99%

https://teslanorth.com/2022/03/09/te...batteries-pic/

Absolutely agree with Andy. thats why i have absorption at 14.2V and 0 minute...my BMS cuts end of charge if a cell is >3,53V constantly for 3 sec, thats 3.57-3.59V directly after cut off for the highest cell.

I tried Jedis Float at 13,15V which is acutally 13.18 at the Bank (voltage drop of shunts,fuses cables) and must say its actually works well as the cells are under load means you are actually still in the >90% SoC range when the MPPT starts to take over the load and then switches to Bulk again. Keeps the battery on a high charge state but low voltage level. Will leave it like this for a while and see longtherm effects.

How good is actually the battery monitor in the Multiplus if you account for the voltage drop of cable and fuses?

[s/v Jedi]

Originally Posted by CaptainRivet

Absolutely agree with Andy. thats why i have absorption at 14.2V and 0 minute...my BMS cuts end of charge if a cell is >3,53V constantly for 3 sec, thats 3.57-3.59V directly after cut off for the highest cell.

I tried Jedis Float at 13,15V which is acutally 13.18 at the Bank (voltage drop of shunts,fuses cables) and must say its actually works well as the cells are under load means you are actually still in the >90% SoC range when the MPPT starts to take over the load and then switches to Bulk again. Keeps the battery on a high charge state but low voltage level. Will leave it like this for a while and see longtherm effects.

How good is actually the battery monitor in the Multiplus if you account for the voltage drop of cable and fuses?

Victron BMV or SmartShunt correctly configured as explained many times now. Maybe, -maybe- some BMS’s can reach equal accuracy but certainly not the monitor in the Multiplus. The manual is clear about that as well.

It is using different gear or mis configured gear that leads to people having trouble.

Take this “you need to use the knee as reference and can’t use SOC” cry: wth is this based on? Certainly not on what I recommend! Maybe some weird $4.- monitor thing from Ali Express?

If you charge for the knee, which my settings do, then you choose a maximum SOC for that voltage at your charge rate in your installation. That’s what it is. Now, if you can’t live with calling that fully charged or 100% SOC then what hope for logic is left? Not much

Err on the wrong side? No, of course not! Not going to the absolute 100% cell soc is the right side to be. Also, a dead accurate read-out for SOC% every cycle is less desirable than one that deteriorates 6% per cycle?

When manuals are written and users do not follow instructions (even when they know they don’t) which leads to errors and next they cry that the results are wrong? Maybe a good old voltmeter would fit better, with one of those sticker labels next to it translating voltage to SOC?

Anyway, can bring horses to the water but can’t make them drink. Cheers!

[JustMurph]

I agree that the Nordkyn design article is excellent and explains things well. I also agree with sailorboy, that no one had cited studies that answer the questions that are actually relevant to us.

For example, the study on over charging to 105 or 110% isn't relevant....none of us are charging that far.

What I want to know, and what I think is most relevant to this group is, if we overcharge by either holding the batteries at an absorption voltage, or a float voltage that is higher than recommended, how much does that impact the life of our batteries.

Essentially, all of these articles are saying that it's bad, but none of them quantify it, right? None of them give examples of failure due to the mild overcharging scenario that's relevant, only extreme examples.

I have an all Victron system which is charged as per Victron's recommendations. What I want to know is, if I swapped from this charging regime to what's recommended by Erik, how much does my battery life increase by? Do I go from 3000 cycles to 4000 cycles? 6000? That's what I want to know. Keen to hear if anyone has that data, either studies or real life experience.

[barcoMeCasa]

Originally Posted by JustMurph

I agree that the Nordkyn design article is excellent and explains things well. I also agree with sailorboy, that no one had cited studies that answer the questions that are actually relevant to us.

For example, the study on over charging to 105 or 110% isn't relevant....none of us are charging that far.

What I want to know, and what I think is most relevant to this group is, if we overcharge by either holding the batteries at an absorption voltage, or a float voltage that is higher than recommended, how much does that impact the life of our batteries.

Essentially, all of these articles are saying that it's bad, but none of them quantify it, right? None of them give examples of failure due to the mild overcharging scenario that's relevant, only extreme examples.

I have an all Victron system which is charged as per Victron's recommendations. What I want to know is, if I swapped from this charging regime to what's recommended by Erik, how much does my battery life increase by? Do I go from 3000 cycles to 4000 cycles? 6000? That's what I want to know. Keen to hear if anyone has that data, either studies or real life experience.

Agree, never seen any major studies about "overcharge by either holding the batteries at an absorption voltage, or a float voltage that is higher than recommended, how much does that impact the life of our batteries."

But I'm with Rod Collins. Why bother?

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

"I have nearly 150 research/white papers in the CMI data-base and not a single one of them gives any good reason to push these cells into the upper knee with regularity. Not a single one."

"If you can get 98% – 99.9% of the capacity out of the bank at a 13.8V – 14.2V maximum charge voltage, why go any higher?"

ISTR his LiFePo4 capacity still being way up near new after a dozen years constant use but can't find a link today.

Anyway, your boats, your call.