LiFePo4 Power -- Controlling Alternator and Inverter/Charger

[Inkog]

Originally Posted by john61ct

Maybe you should read some past threads here to become familiar with the community.

The usual turnkey packaged systems here are from the vendors I cited, usually installed by professionals and cost thousands.

The drop-ins you're talking about are widely considered sub-par by those whose opinions are most respected here.

There are hundreds of systems in use where the owner has bought bare cells from a known-good LFP maker - e.g. Winston/Thundersky/Voltronix, CALB, GBS, Sinopoly - then selected an appropriate BMS - as with Batrium as discussed in the OP - and put in a well designed architecture of quality gear far superior to what's inside any drop-in.

DIY does not mean poorly implemented, but does take a decent level of knowledge and skill.

Trojan's Trillium line has recently started shipping, apparently OEM is K2, and their more sophisticated approach, along with Victron's lower end lines, are a sort of "hybrid drop-in", perhaps closer to the top packaged systems.

Of course it will be years before we know whether they measure up or not.

Fair enough. We were speaking of charge current of LiFePo4. I just looked up recommended charge currents for your high-end, non drop-in vendors:

Lithionics - .5c
Ocean Planet - Couldn't find their LifePo battery specs, but **** they are expensive! Their website sucks.
Victron - <.5C
MasterVolt ~.33C

Cell manufactures (with no direct warranty):
Winston - max 3C, no recommended
Volttronix - max 3C, no recommended
CALB - Max 1C, no recommended
Sinopoly - .3C

Doesn't matter, they are all the same chemistry. Fancy management systems can't overcome the limitations of chemistry. If you want to charge LiFePo4 really, really fast, and make them last, just get a bigger bank (or a cooling system)

[john61ct]

Lithionics is the manufacturer, Bruce @ OceanPlanet helped design their packaged system and is their lead reseller / distributor. They are also a Victron dealer.

[Nicholson58]

Originally Posted by Inkog

A battery needs potential to charge. This means that the voltage of the charger has to be higher than the voltage of the battery. If your Li bank has been charged to 100% at 13.8v (resting) and your charger goes into Float mode at 13.6v, the battery cant and wont take a further charge. There is no harm done to the battery if the "float" charge is below the nominal battery voltage.

The only reason you shouldn't pull a lot of amps from your bank while in Float is that any voltage sag from the load could cause the charger to go back into Bulk and/or Absorption which is at higher voltage, so you'd be back to charging your full battery. The key here is that you want a charger that has a highly configurable absorption mode. If you can turn it off or adjust the voltage to be the same as your float voltage (13.6v in this case) that's ideal. Otherwise set the absorption for the shortest time possible. Often this is defined as a fraction of the time spent in bulk.

If you can turn off or configure absorption voltage, and configure float voltage to be below the banks nominal voltage there is no need to micro manage your charging system. Going into Bulk as needed is perfectly fine. You can adjust the bulk voltage down 80%-90% resting for longevity vs max capacity.

I do not mean to argue, but you either did not read or did not understand what was taught in the link I included. If you float a lithium set, you will damage it. A BMS, among other functions is designed to separate the charging sources from the batteries when the fist cell reaches full charge. The guy who wrote the piece is an electrical engineer who built and studied lithium batteries and banks including over charge and float. The conclusions are absolute. You must consider breakers and relays as part of the BMS as well as modified programs to accommodate lithium charging. Most of the makers we deal with for charge control have realized this and now offer program patches or new items with a profile for lithium charging. Do not take this lightly or you will not get the service you paid for. Several of us retired PEs have studied this in depth. Please follow the link and read all including the other embedded links. Note his opinions regarding so called drop in replacements that look like standard batteries.

For our boat, the total of all changes and equipment made us re-think. I got much enhanced and acceptable service with no re programming or wiring using Firefly. Downloading and installing updates in the Caribbean would be problematic. Firefly is a souped up AGM so no changes needed. Plug and play. Several of the other guys based in the US touched their way through the changes with factory help.

[john61ct]

Originally Posted by Nicholson58

Read this.

There is NO FLOAT for Lithium. Float will wreck them.

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

I completely agree with you, and Maine Sail **in principle** on that topic, best to default to being ultra conservative if maximizing longevity is important to you.

However in practice, isolating the LFP bank from energy inputs while varying-current concurrent loads are unpredictably coming online and going off, is a pretty complex challenge for many setups.

There is no evidence either way whether or not

having float continue at below the 90% SoC resting voltage, say 3.28 - 3.32Vpc

will have a significant impact on longevity.

> The conclusions are absolute.

Here is that section, you will find few "absolute" statements

______

FLOAT CHARGING:

These are not lead acid batteries and they were not designed nor intended to be “float charged” in the typical lead acid sense. There is scant data on float charging LFP cells. We have two years of testing float on LFP and unfortunately the data collected is all over the map. While we have a few premium branded LFP cylindrical 18650 cells that suffered zero capacity changes, after being held at 3.400V (13.6V for a 12V nominal bank) 24/7, for six months continuously, we also have capacity losses holding the same voltage on counterfeit cells and no-name LFP cylindrical cells exceeding 16%. We also have an 11% Capacity loss on some CALB SE prismatic cells from leaving them at 100% SOC and letting them sit for 12 months. In other-words the data is confusing at best and our float life testing here is still on-going.

Look at any of your tablets, cell phones iDevices etc. and they all terminate charge when the battery is full. They cut back in when battery terminal voltage has fallen to a preset level, but they do not hold a high voltage on a full battery.

Floating LFP is a certainly a complex subject with scant data. Bottom line is to avoid floating LFP banks if you can. If you absolutely must do this you would be best to be below 13.6V or 3.400VPC. Some have argued that a float voltage of 3.35VPC or lower (13.4V for a 12V nominal bank) is not badly damaging. I would simply counter that with we don’t have enough data, across all cells, to confirm that. There is very little research and literature on floating LiFePo4. This type of charging, if the voltage is high enough, keeps you in the upper SOC range for long periods of time and these batteries, according to every LFP cell maker we know of, prefer to sit at a mid-range SOC when not being used. These cells were originally designed to be actively cycled.

Can you float at 3.400VPC or 3.35VPC or lower? Sure, you can do what ever you want to, but we don’t really know the long term affects other than to say it is it may shorten the life of some cells and may cause no harm to others. The premium cylindrical cells we tested at 3.400VPC (using a very expensive very linear power supply, lost no quantifiable capacity but some of the cheap cells lost as much as 16% in the same time frame. Do you or will you know the quality of the cells inside your own battery and how they actually handle being floated?

Premium Cell #1 – 1100 mAh Rated – Base Line Capacity = 1.140 mAh – 6 Months at 3.400VPC = 1.130 mAh

Premium Cell #2 – 1500 mAh Rated – Base Line Capacity = 1.391 mAh – 6 Months at 3.400VPC = 1.387 mAh

Premium Cell #3 – 1500 mAh Rated – Base Line Capacity = 1.404 mAh – 6 Months at 3.400VPC = 1.403 mAh

No-Name Cell #4 – 1200 mAh Rated – Base Line Capacity = 1.101 mAh – 6 Months at 3.400VPC = 0.921 mAh

Counterfeit Cell #5 – 1500 mAh Rated – Base Line Capacity = 1.298 mAh – 6 Months at 3.400VPC = 1.192 mAh

If you will note above that only one of these cells delivered it’s rated capacity, cell #1. Three of them, the premium branded cells, lost virtually no capacity after six months at 3.400V and the two other cells, no-name brand and a counterfeit of one of the branded cells, lost quite a bit of usable capacity when floated at 3.400V or 13.6V for a 12V bank.

Of the piles of white papers I have on LFP batteries not a single one of them has dealt with fractional “C” use of LiFePO4 and floating at 3.400VPC. The only float paper I have was using Mn doped LiFePO4 cells and in 24 months cells floated / maintained at 100% SOC and at 25ºC lost 30% of their capacity, without any cycling. What we do know is that storing these batteries at 100% SOC resting voltage (not even charging) can lead to capacity loss and is advised against by every LFP cell maker I know of.

The question of floating LFP, and its impact on cycle life, is still very much unclear.

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

[john61ct]

Originally Posted by Nicholson58

A BMS, among other functions is designed to separate the charging sources from the batteries when the fist cell reaches full charge.

Well there are many types of balancing circuits, but most of the "bleeding resistive" type you describe just do that long enough to burn off the topmost cells, then resume charging them altogether.

The BMS HVC should not be relied on for charge termination.

> The guy who wrote the piece is an electrical engineer

Do you mean Maine Sail there?

[Inkog]

Originally Posted by Nicholson58

I do not mean to argue, but you either did not read or did not understand what was taught in the link I included. If you float a lithium set, you will damage it. A BMS, among other functions is designed to separate the charging sources from the batteries when the fist cell reaches full charge. The guy who wrote the piece is an electrical engineer who built and studied lithium batteries and banks including over charge and float. The conclusions are absolute. You must consider breakers and relays as part of the BMS as well as modified programs to accommodate lithium charging. Most of the makers we deal with for charge control have realized this and now offer program patches or new items with a profile for lithium charging. Do not take this lightly or you will not get the service you paid for. Several of us retired PEs have studied this in depth. Please follow the link and read all including the other embedded links. Note his opinions regarding so called drop in replacements that look like standard batteries.

For our boat, the total of all changes and equipment made us re-think. I got much enhanced and acceptable service with no re programming or wiring using Firefly. Downloading and installing updates in the Caribbean would be problematic. Firefly is a souped up AGM so no changes needed. Plug and play. Several of the other guys based in the US touched their way through the changes with factory help.

I can only assume you're speaking of the fact that LiFePo4 doesn't like to be maintained at 100% charge for long periods of time. This is true, you shouldn't use a LiFePo4 bank in an environment where it is "floated" at near 100% SOC constantly. LiFePo4 is not a battery designed for UPS service. I assumed in this context we where speaking of applications where we'd be using these expensive batteries for their intended use - Longevity through deep discharge. This conversation doesn't make sense if you aren't deeply discharging your batteries regularly. Therefore neither does your argument. This is the whole point of using LiFePo4 over lead acid.

If you need to store your batteries for some time, turn off your charger and drain them to 50%.

[Inkog]

One day someone will make a dedicated LiFePo4 charger specifically for our needs.

I bet it will have an equivalent "float" profile intended to supply excess power to loads while keeping the bank at an ideal SOC.

Maybe no one has because you can do the same thing with many chargers already available?

[john61ct]

Originally Posted by Inkog

One day someone will make a dedicated LiFePo4 charger specifically for our needs.

I bet it will have an equivalent "float" profile intended to supply excess power to loads while keeping the bank at an ideal SOC.

Maybe no one has because you can do the same thing with many chargers already available?

SoC ideal for what goal?

The ideal is keep the bank at a low SoC, by default, for most of its life,

only start charging it up just before it will be required to service loads, i.e. when the other energy inputs will no longer be available.

Especially when there are 2-3 or more available charge sources, traditionally not communicating with each other at all,

that's a pretty big challenge to automate.

It is the desire to have everything run without the owner having to monitor & control that makes these finer points more challenging.

If you know your setup and babysit manually, becomes much easier.

Of course last-resort HVC and LVC and temperature cutoffs are required for failsafe protection.

But as many have noted "a BMS" as such is not a panacea, and many do without them completely.

[Inkog]

Originally Posted by john61ct

SoC ideal for what goal?

The ideal is keep the bank at a low SoC, by default, for most of its life,

only start charging it up just before it will be required to service loads, i.e. when the other energy inputs will no longer be available.

Especially when there are 2-3 or more available charge sources, traditionally not communicating with each other at all,

that's a pretty big challenge to automate.

It is the desire to have everything run without the owner having to monitor & control that makes these finer points more challenging.

If you know your setup and babysit manually, becomes much easier.

Of course last-resort HVC and LVC and temperature cutoffs are required for failsafe protection.

But as many have noted "a BMS" as such is not a panacea, and many do without them completely.

The great thing about LiFePo4 is their higher voltage throughout their discharge profile. A 50% SOC LiFePo4 is still well over 13V. You can "float" your system at 13v forever if you need. All your 12v systems wont care in the slightest. Of course this is at the detriment of capacity.

90% SOC is fine if your trying to squeeze the most life out of your bank.

Multiple charge sources don't really matter if their all configured correctly. Resistance determines bulk charging. All chargers are designed around this and our LiFePos are no different. Its Absorption that we have to worry about. Lead acid needs to be forced into taking their last ~15% (or more as they age), where our LiFePos don't, or at least shouldn't. That 15+% is big when you can only use <50%, and why all lead acid chargers have that stage. If you can make sure all of your chargers skip this stage or drop their voltage during it doesn't matter (you need potential, or higher voltage to charge). We've been over Float...

You can have 200 chargers bulk charging your bank and they'll all taper the amperage as the battery resistance dictates. That's how all batteries work. LiFePo will be about 95% charged at this stage and we can/should quit here. We're now at ~13.8+ volts resting. If we "float" we may drop to the float voltage if their is loads on the system. If we have a big load like running an inverter, we'll drop resistance quickly. The charger will see the drop in resistance and the process starts all over.

No need to automate. No need for a BMS to dictate charge, most don't.

[Q Xopa]

Originally Posted by john61ct

Well there are many types of balancing circuits, but most of the "bleeding resistive" type you describe just do that long enough to burn off the topmost cells, then resume charging them altogether.

The BMS HVC should not be relied on for charge termination.

> The guy who wrote the piece is an electrical engineer

Do you mean Maine Sail there?

The previous poster was referring to not 'Floating'.
You are now talking about 'balancing'?
I am not sure what your point is?

[john61ct]

The post I was quoting referred to a behaviour that they thought was related to BMS control of charge termination.

I was pointing out it was actually related to the balancing process.

[Q Xopa]

Originally Posted by john61ct

The post I was quoting referred to a behaviour that they thought was related to BMS control of charge termination.

I was pointing out it was actually related to the balancing process.

Ah, ok I see. Thanks I've learnt something new. I never considered BMS balancing and Charge termination were so related.

[john61ct]

They don't have to be.

There are dedicated balancing-only gadgets ("non-protective" BMS) that can keep balancing at any / all SoC.

Most shunt current from higher voltage cell/groups to the lower voltage ones, rather than resistive-bleeding like mainstream ones.

And at much higher balance-current rates, rather than IMO stupid-slow fractions of an amp.

There are also protective BMS where the user can adjust the start-balance voltage setpoint much lower, say 3.4Vpc, so the balancing gets a head start much earlier in the charge cycle's ending stage.

I personally think those that don't even start before getting to 3.6Vpc or whatever, are ridiculous, since many owners don't want to go that high at all, ever in normal cycling.

Then having a rate of say 50mA adds insult to injury, if you happen to have a lower-quality, or a mixed bag of poorly matched cells, or aged ones, it can literally take many **days** of sitting at that over-high voltage in order to finish balancing.

Which often means the job never gets done, since most owners have no idea about these details.

But of course, a new bank composed of top-quality matched cells, in both capacity and internal resistance, may if high C-rates and the voltage shoulders are avoided, may never need balancing.

In which case all these concerns are moot.

Until the cells age and start losing capacity, as we all do eventually.

[Q Xopa]

My fault. Any chance we can get back on topic.

[tanglewood]

Originally Posted by Nicholson58

I do not mean to argue, but you either did not read or did not understand what was taught in the link I included. If you float a lithium set, you will damage it. A BMS, among other functions is designed to separate the charging sources from the batteries when the fist cell reaches full charge. The guy who wrote the piece is an electrical engineer who built and studied lithium batteries and banks including over charge and float. The conclusions are absolute. You must consider breakers and relays as part of the BMS as well as modified programs to accommodate lithium charging. Most of the makers we deal with for charge control have realized this and now offer program patches or new items with a profile for lithium charging. Do not take this lightly or you will not get the service you paid for. Several of us retired PEs have studied this in depth. Please follow the link and read all including the other embedded links. Note his opinions regarding so called drop in replacements that look like standard batteries.



For our boat, the total of all changes and equipment made us re-think. I got much enhanced and acceptable service with no re programming or wiring using Firefly. Downloading and installing updates in the Caribbean would be problematic. Firefly is a souped up AGM so no changes needed. Plug and play. Several of the other guys based in the US touched their way through the changes with factory help.

There is a terminology assumption in MaineSail’s comments of floating LFP that unfortunately is leading people to the same incorrect, or at least incomplete conclusion about float.

“Float” in the context of LA is at an elevated voltage that causes an ongoing charge current to the battery. You don’t want to do that with LFP, and that’s what MaineSail cautions against.

But you can still safely use the “float” function in many chargers to carry loads rather than disconnecting and gratuitously cycling LFP batteries when a power source is available. The key is to set the float voltage such that there is no ongoing charge current to the batteries. This is done by matching the float voltage to the battery OC voltage at whatever SOC you elect to
maintain. 70-80% seems a common choice.

This later part is what you are missing, and what MaineSail unfortunately doesn’t address in his article. So everyone is blindly following this “don’t float” mantra without really understanding the issue.

The mantra should be “stop charging”, not “don’t float”, where “charging”
Is defined as a current flowing into the battery.