LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[T1 Terry]

Well, I'm off for a mth travels, see you all when I return. Be good, no fighting or back biting now :lol:

T1 Terry

[deckofficer]

mth? month or motorhome? where you going? We'll be good.

[ebaugh]

Originally Posted by T1 Terry

Some understanding and application of that understanding is needed. A cell is 100% full with a rested voltage of 3.4v, you can't have 105% or 110% full can you? There is no added capacity above 3.4v rested, but charging a to 3.4v will not achieve a 3.4v rested voltage reading, think about why this would happen and you are part way to understanding lithium ferrous battery charging requirement, particularly if a return to full capacity is required.
T1 Terry

There is an important factor missing. Charge rate at termination of charge. Charging to 3.4V and stopping at .5C charge rate is very different than charging to 3.4V and stopping at .005C.

I just left port with a battery around 70%. 48 hours later, never charging beyond 3.35V, my 1200Ah bank was no more than 50Ah from full charge. How do you explain that? It's a fact. It was unexpectedly higher than I anticipated, that's why I keep mentioning it. But it assures me that if I had made the "float" voltage 3.4, not 3.35 they would have been exceptionally close to fully charged.

It looks to me like if you hold a specific voltage long enough, the cells will charge to whatever that level represents. Suppose you charge one cell normally to 3.6V and stop, leave another similar cell at 3.4V until there was no more than a few milliamperes of charge current, they would have essentially the same capacity. I will offer the idea there are only so many electrons to push around and either way they all get pushed over.

This is why if more than 3.4V is held too long, overcharging could occur. The question remains, what value is optimal. You either must float your LFP pack somewhere when excess charging capacity is available, or remove it from the system until needed. No one has proposed a system to remove it for marine.

[ebaugh]

Originally Posted by goboatingnow

I dont think that ANY float charge should be applied to an LI battery set, All my reading has shown that Li is damaged ( shorter life) by applying ANY induced voltage when the battery is already full.

Where that cant be done, ie where you have load sharing ( between battery charger and teh battery), I would reccomend that the battery( cell) never be charged above about 80%, and then floated at 12.8 (3.2v) , and certainly not at 3.6v per cell.

Then at least the charge ( voltage stress) on Li is reduced at 80%, and the battery can handle the float charge.

Far better to stop charging and then only recharge when a lower voltage threshold is triggered. There is no advantage to float charging Li, as the self discharge is incredibly low.

What this brings up is actually the unsuitability of Li for applications involving load sharing. Im grappling with this at the moment in Li based battery backup product Im designing . The quickest way to kill Li, is to leave it on high float for any length of time. Li is great for "charge, discharge , recharge " applications, like drill batteries, phones, laptops. It sucks at load sharing , because you are applying a charge voltage continuously. The only way around that, other then experiencing shorter cycles, is to charge to about 80% , ie very conservative cutoffs.

Its a funny irony , that while we are gaining theoretical additional discharge capacity , the restrictions on various issues with Li , mean in practice we gain about 60% of teh batteries capacity ( ie 80% ->20%) , whereas with LA we have nominally 50%.

If you want high capacity charging , ie to 95-100% , then I would not reccomend any load sharing configurations ( ie when fully charged)

Dave

Dave,

Can you quote some reading sources here? I've not been able to find any that support "floating" is bad, just "float charging".

I did find an EV guy who floated some LFP cylindrical cells for 6 months at 3.45V and was not able to measure any capacity loss whatsoever. I'm a bit surprised he did not overcharge them some at that level, but if so, it did not do enough harm to impact "as new" capacity. I will grant it does not prove there was no impact on cycle life, but generally LFP experience an almost linear degradation in cycle capacity during cycle testing. So I think it can be inferred, there was no cycle life lost either.

I know little about Li chemistry outside of LFP, other than what I've been unable to avoid reading when looking for LFP specific information. Is it possible LFP behaves different?

I would argue the easily usable range is 10-90%, but the 10% on the bottom does not leave the same fudge factor available on LA with a 50% bottom.

[ebaugh]

Originally Posted by hellosailor

And the shift to Yttrium doping (which Sinopoly seems to be implementing across only portions of their line) makes no difference in the charging recommendations, just in the performance?

I think the Yttrium and Mn, specifically LiFePO4 vs LiFeYPO4 vs LiFeMnPO4 are marketing distinctions. And impose no practical operational differences in regards to voltage. The unsubstantiated arguments I've seen are adding "Y" could increase cycle life and the "Mn" could increase cold weather performance. But even those references are obscure. So due to the lack of data...it's marketing until there is data to demonstrate the distinctions.

[diugo]

Originally Posted by ebaugh

I did find an EV guy who floated some LFP cylindrical cells for 6 months at 3.45V and was not able to measure any capacity loss whatsoever. I'm a bit surprised he did not overcharge them some at that level, but if so, it did not do enough harm to impact "as new" capacity.

There's a handful of anecdotal evidence on the Web that LFP current "drops to zero" under protracted constant voltage, and both TerryT1 and MaineSail have apparently observed this as well.

However, every one of them may have been using a Switched Mode Power Supply or a charger that employs SMPS technology. Under SMPS, output voltage is monitored and output current is reduced to zero as the target voltage is approached. Essentially, the charger exploits the LFP's extremely low self-discharge characteristic and eventually turns itself off---providing only a periodic topping charge.

A much different effect might be observed with a less sophisticated charging circuit, such a simple transformer-rectifier combo. There, output current is just a function of the internal resistance presented by the battery. As the LFP enters the upper knee, resistance increases and current drops until a (nonzero) current equilibrium is attained, eventually resulting in overcharge.

[ebaugh]

Originally Posted by diugo

There's a handful of anecdotal evidence on the Web that LFP current "drops to zero" under protracted constant voltage, and both TerryT1 and MaineSail have apparently observed this as well.

However, every one of them may have been using a Switched Mode Power Supply or a charger that employs SMPS technology. Under SMPS, output voltage is monitored and output current is reduced to zero as the target voltage is approached. Essentially, the charger exploits the LFP's extremely low self-discharge characteristic and eventually turns itself off---providing only a periodic topping charge.

A much different effect might be observed with a less sophisticated charging circuit, such a simple transformer-rectifier combo. There, output current is just a function of the internal resistance presented by the battery. As the LFP enters the upper knee, resistance increases and current drops until a (nonzero) current equilibrium is attained, eventually resulting in overcharge.

All modern chargers have voltage regulation. They have to in order to handle different programmable charge profiles and the temperature adjustments for LA. I'm not sure I see a difference between those using SMPS or transformers?

Many years ago, I built a couple of home brew power supplies for communications equipment. I never got it fully right since it was a transformer, bridge rectifier and a couple of capacitors. It worked, but the RX voltage was very different than the TX voltage because of the current draw. But let's say I built another one, where the "no load" voltage was 3.4V and hooked it up to a LFP cell. Are you saying somehow it would never stop charging? Or that it would somehow never reach 3.4V, but continue to charge indefinitely at some lower voltage?

It would be different if you add random loads since they would be powered by the battery, where modern chargers increase current to hold the voltage constant, and they are pretty snappy about it. Is this the difference you are highlighting?

[FlyingCloud1937]

Magnum Inverters have a 4th stage called battery saver

This works very well and would be applicable to LI-charging.

Full Charge (Battery Saver™ mode):
The fourth stage occurs after four hours in the Float Charging stage. The Full Charge stage is designed to keep batteries fully charged over long periods, and to prevent excessive loss of water in flooded batteries or drying out of GEL/AGM batteries. In this stage, the charger is turned off and begins monitoring the battery voltage; if the battery voltage drops low (12.7 VDC or less on 12-volt models or 25.4 VDC or less on 24-volt models), the charger automatically initiates another four hours in float charge.

Info:
If the battery voltage falls to the re-bulk voltage (12.1 VDC on 12-volt models or 24.2 VDC on 24-volt models) or lower, the unit will begin another bulk charge.

This allows the the voltage to fall to zero if no loads exist, and quickly move back into power supply mode, when a load comes on line. So no load from the battery when in stand-by mode.

Lloyd

[ebaugh]

Originally Posted by FlyingCloud1937

Magnum Inverters have a 4th stage called battery saver

This allows the the voltage to fall to zero if no loads exist, and quickly move back into power supply mode, when a load comes on line. So no load from the battery when in stand-by mode.

Lloyd

I have a Magnum, this feature does exist, but it's not configurable or defeatable on my unit/software level. With LFP it simply lets the battery carry the full house load to 12.7 or 12.8 V, lower than I'd like to begin a recharge. That's going to be 15-25% SOC on LFP before it charges again.

I don't think it's needed, but if you desire this feature, the voltage levels need to be changeable to shallow the discharge.

[FlyingCloud1937]

Originally Posted by ebaugh

I have a Magnum, this feature does exist, but it's not configurable or defeatable on my unit/software level. With LFP it simply lets the battery carry the full house load to 12.7 or 12.8 V, lower than I'd like to begin a recharge. That's going to be 15-25% SOC on LFP before it charges again.

I don't think it's needed, but if you desire this feature, the voltage levels need to be changeable to shallow the discharge.

What remote are you using? Also I think it's possible to upgrade the inverter software, call tech or email support at Magnum.

With my unit i can set each parameter.

Lloyd

[ebaugh]

Originally Posted by FlyingCloud1937

What remote are you using? Also I think it's possible to upgrade the inverter software, call tech or email support at Magnum.

With my unit i can set each parameter.

Lloyd

At the time I first saw this, I was only interested in defeating the logic since the values were not adjustable or useful for me. With the help of Magnum support (very good!), we determined I needed both the newest remote controller (even though it looks identical) AND a firmware upgrade on the board in my inverter charger.

Sounds like now it will let me not only adjust, but defeat the logic. I seem to remember the tech saying the voltage levels for Battery Saver were hard coded, unlike the other customizable values. Maybe they have fixed that in the last 9 months or so.

[hellosailor]

"This link (from a CF post back around 1250) clearly shows that a LFP prismatic "cell" is actually a "battery" made from many flat pouch cells in parallel."

One could be generous and call that a translation error made by someone too cheap to hire an engineer and technical copywriter to check the translations. But technically, a cell is a cell and those are batteries, which makes calling them "cells" a simple fraud. Yes, it is that simple.

That's more trust and respect I'm not having for these guys.
"The Moe, Larry, & Curly Battery Company" ?

Really? How bleeping hard IS it to just be honest about what the product is?

[deckofficer]

How can you knock it? When a lead acid battery maker gives an a-hr rating, it is based on 20 hours. As you have seen by Terry, Maine Sail, myself and others, the LiFePO4 are rated very conservative, and at the 20 hour rate give about 25% more than their rating. Now that is truth in advertising because the a-hr rate is what we go by in sizing our banks. Not to mention almost nil Peukert effect, and 80% DOD vs 50% DOD, I would say the LiFePO4 makers are being more honest about their battery than the LA group.

[rainmaking]

Originally Posted by roetter

Great maybe this thread can get back on topic, and stop the doomsday predictions. I don't see/hear anyone twisting anyone's arm to dive into this. If you don't want to tread here, fine, but please move on.
I know nothing of all of this, oh yea my 700 ah hour cells arrived a couple of weeks ago (on time and as advertised from that Bankrupted Co.) and so far all I know is you don't want to experience a complete internal short, however, compared to the equivalent LA event, quite tame and safe, really! (PS: My fault not the batteries)

How did the shipping go? I am looking at the 700Ah ones for two boats. I need them to go to Florida.[/QUOTE]

Roetter,
I am in SFL and am ordering 12x 700Ah SP's. This thread is great but hard to weed through a lot of the crap people want to stir up. I am tuned in to all the issues here, and even have some new ones (like non-adjustable outboard motor alternators). I really appreciate ebaugh's recent posts related to real and relevant experience 'float' charging at 3.35V.

[roetter]

A cell is still a cell even with multiple pouches.

For high capacity (Ah) you need surface area. If you do not want to have a cell that is several feet wide, than you need to stack plates on top of each other. In Lithium you do that in several pouches, placed in parallel in one cell. It still remains a cell.

A battery has several cells in series to achieve higher voltages.

In LA batteries you also have multiple plates. In parallel in a single cell and feed them to the same terminal. So what is the difference? The only difference it that the LA plates "swim" in the same chamber in Lithium they are in hermetically sealed pouches. That is just done for convenience of manufacturing. They remain cells.