What is "real" truth about IFR (LiFePO4) cells?

[Rosoku Chikara]

I have read both of these two extremes on this forum:

1) IFR cells are the perfect solution to all Li-ion "worries."
"Just use IFR and your will never need to check voltage, or worry about undesirable events such as fires, etc., ever again." <paraphrased from memory>

2) The chemistry of IFR cells is different, but they are no safer than other Li-ion cells.
"The differences in chemistry and construction do solve (minimize?) issues pertaining to one pole of the cell (+ or -, I have forgotten which), but the same risks as other Li-ion cells remain on the opposite pole, so in reality, IFR cells are not really any safer at all." <paraphrased from memory>

I suspect that the truth may lie somewhere in between. Will some of you with better understanding and experience, please provide me with some practical ("real world") advice?

Are IFR cells any "safer" when being used in a typical flashlight and/or during recharging?

Looking forward to any and all comments and opinions.

 

[AnAppleSnail]

LiFePO4 cells are more forgiving of cell failure. When things go wrong, these chemistries tend to fail in the less exciting ways. However, a dead short of a charged 20Ah 12v pack will be exciting no matter what chemistry is used. Protection from that requires a battery to integrate several safety mechanisms (PTCs, fuses, physical isolation). That's beyond the chemistry.

LiFePO4 has less of a 'thermal runaway' effect than a LiCo battery (The famous "Exploding Laptop" batteries were LiCo or LiMn with metal shavings causing dead shorts). But "Never checking voltage again" is a fast road to a pack with one dead cell and several healthy ones. They might be more tolerant of dead-drain conditions, but I may remember wrongly.

In other words, you can get away with more, but a dead cell will die from the same neglect. And they can still cause problems - Any high-density energy storage can be harmful.

 

[jtr1962]

Based on my experience, I think LiFePO4 brings two things to the table-relative freedom from dramatic failure modes, and much longer cycle/calender life. Any type of rechargeable battery can cause issues if it's shorted, ran at higher than rated current, overcharged, or reverse discharged. However, with LiFePO4 those things cause the same issues they cause with NiMH and NiCd. The cell may not work as well or at all after such an event, but it will still generally be intact (perhaps with a melted wrapper). This is quite unlike the other lithium chemistries which have a propensity to explode if abused.

The second thing is the one I consider more important. Other lithium chemistries last 300 to 500 cycles, or 3 to 7 years, whichever comes first. LiFePO4 typically last at least 1000 cycles, although most are rated for upwards of 2000. As far as calender life, the jury is still out, but I have some A123 26650s from ~2006 which perform like new. It may well turn out that LiFePO4 have a calender life similar to that of NiCds-perhaps upwards of 20 years, maybe even more. For projects where longevity is more important than absolute capacity, LiFePO4 is the best choice.

The major disadvantage of LiFePO4 is lower energy density. For small cells like 18650s energy density is about half of other lithium chemistries. As cell sizes get larger, the energy density disadvantage only drops to maybe 20% to 30%.

 

[HKJ]

It is not that simple.
LiFePO4 has lower voltage, that means they will not work with lights designed for single cell LiIon, but they can sometimes be used for lights designed for single CR123.
LiFePO4 does not have a protection circuit, this can be a problem if they are shorted or run dry.
Batteries like A123 are nasty when they are shorted, due to the extreme high current, the battery might not burn, but the wire doing the shorting probably will.

LiCoO2 is often safe today, they have been tested with shorts and overcharge and they must handle this without a protection circuit (Handle does not mean survive, but only no fire or explosion). With a protection circuit they can handle it without any serious damage.


The main problem with old style LiCoO2 is that they goes into thermal runaway at a lower temperature than LiFePO4. In many modern batteries this is fixed with a special layer inside the battery, that will turn the battery off, if it gets to hot.

 

[Nisei]

And then there's also this.

 

[StorminMatt]

To better give you an idea of how tough LiFePO4 is, let me tell you what I have done to LiFePO4 batteries and lived through:

1. I accidently put an A123 Systems 26650 in the charger with the voltage setting on 3.7V. The battery ended up charging to 4.1V. But it did not get warm or show any signs of impending problems. The battery seems to have suffered no ill effects from this overcharging.

2. I overdischarged an A123 Systems 18650 to .4V and left it there for an unknown time (perhaps a week or two). When I went to charge it, the charger refused to do so. That's when I discovered the low voltage. So I revived it the same way you would a dead Eneloop by putting it in parallel with a good one and a Mag bulb between the two for current limiting resistance. To this day, the battery is fine.

3. I discharged some Tenergy 14500s at MUCH higher than the rated current. For some reason, Tenergy says the maximum draw is 1C, of 400mA. I used two to power a ROP Lo bulb, which is a 2A draw. Not only did the batteries handle it fine. But the voltage drop across the two batteries wasn't too bad (maybe .2V). Maybe Tenergy underrates these. But I drew several times the maximum current.

4. I have had Tenergy RCR123A batteries go through the wash with no ill effect. Of course, LiCo could possibly survive as well - as long as it doesn't have a protection circuit that can potentially be damaged by water.

 

[Rosoku Chikara]

Thanks for all the good posts. Based on these comments, I guess my final conclusion is that while IFR is no "silver bullet," IFR cells are indeed significantly "safer." And, to the extent that they are safer, they probably are a better cell for me.

Not to say that I am fearful of potentially dangerous activities or products. I have used, and currently use, many items that if mishandled would be extremely dangerous. But, I have never felt the need for Li-ion cells. I own several types of Li-ion cells, and use them for testing my flashlights; just to see how much brighter they truly are on Li-ion. But, I purchase only AAA and AA lights, so I have always found Eneloops to be the most suitable for my day-to-day needs.

But, now, I am being tempted to start using IFR on a daily basis.

Let me verify one last point: I gather that vigorous or even semi-vigorous "out-gassing" is still a potential issue for IFR cells? So, in the unlikely event that such an unfortunate thing happened to a IFR cell contained in a waterproof flashlight, I presume I can still expect that flashlight could "go off like a pipe bomb," unless it had some kind of "weak link" (intentional, or unintentional pressure relief system) built into the pressure container (waterproof flashlight body)?

Thanks again for all your comments.

 

[cy]

adoption for 12v LiFePO4 batteries for motorcycles is happening at a dramatic rate. LiFePO4 has many advantages over other Li-ion chemistries. the world's internal combustion vehicles runs off 12v or 6v or multiples of .. just so happens charging systems for ALL 12v PB matches perfectly with ALL 12v LiFePO4 batteries. with NO modifications needed to run 12v LiFePO4 batteries .. this means all 12v LiFePO4 batteries are drop-in with no mods needed. but there's always a catch .. that's provided AH capacity of 12v LiFePO4 battery installed is the correct size.

while 12v LiFePO4 seems far from flashlights .. characteristics for single cell LiFePO4 or multi cell remain the same. as it's already been noted above.. as cell size goes up energy density disadvantages compared to LiCo chemistries narrows. LOTS more information inside link below

if you are searching for LiFePO4 info .. my LiFePO4 testing thread has morphed into world's largest database for 12v LiFePO4 motorcycle batteries http://www.advrider.com/forums/showthread.php?t=757934

 

[Viking]

Hi Cy

In your link you have listed a very interesting state of charge table for the lifepo4 chemistry.
Will it provide an equally precise indication for 3.2 v batteries if I divide the 6 volt column in 2 ?

7.17 volt/2 = 3.585 volt = 100%
That corresponds very well with the numbers I get , when I measure the cells fresh from the charger.
Only problem , they will drop down almost immediately to about 3.34 volt , where they will rest forever if not used.
That corresponds to just above 90 % according to your table. That is if I can divide the 6 volt column in two of course.
Is that correct , will they loose that much capacity that quickly ?


But aside from that , I find all the other capacity numbers very likely according to my personal experiences , if the 6 volt column is dived in two.

But can I do that. Will it be an equally precise ( roughly of course ) indication ?

 

[StorminMatt]

Hi Cy

In your link you have listed a very interesting state of charge table for the lifepo4 chemistry.
Will it provide an equally precise indication for 3.2 v batteries if I divide the 6 volt column in 2 ?

7.17 volt/2 = 3.585 volt = 100%
That corresponds very well with the numbers I get , when I measure the cells fresh from the charger.
Only problem , they will drop down almost immediately to about 3.34 volt , where they will rest forever if not used.
That corresponds to just above 90 % according to your table. That is if I can divide the 6 volt column in two of course.
Is that correct , will they loose that much capacity that quickly ?


But aside from that , I find all the other capacity numbers very likely according to my personal experiences , if the 6 volt column is dived in two.

But can I do that. Will it be an equally precise ( roughly of course ) indication ?

I don't know that this drop in voltage necessarily represents a significant loss of capacity. From what I've seen, it seems to be similar to what happens with NiMH if you 'rest' it after charging. You will experience a drop in the open circuit voltage of the cell. But that doesn't mean the cell has suddenly lost significang capacity compared to one that was used without resting. Also remember that, with LiFePO4, this drop from float to plateau voltage occurs much more quickly than it occurs with NiMH. This suggests that any amount of energy that could be lost when a battery rests is negligible.

 

[BVH]

LiFeP04 has another yet mentioned benefit over LiPo. It can be stored/not used for very long periods of time at full charge. Lipo's suffer significant damage - shorter life span, when stored at full charge. So much so that the RC people in-the-know, won't let their Lipos sit for more than a day or two at full charge before they discharge them down to about 3.85 Volts per cell. Any of the "better" hobby chargers have specific routines to read cell voltage and then automatically initiate a charge or discharge to set the cells to the storage voltage.

Another advantage for me anyway, is the very flat discharge curve. For powering heavy current 28 Volt nominal searchlights, I can get only about 40% of the rated capacity out of the pack before Voltage falls below acceptable whereas I can get 80 to 90% capacity out of a pack with LiFeP04 before Voltage falls too low. You can't simply add another cell to the Lipo pack because you would fry the electronics.

 

[cy]

Hi Cy

In your link you have listed a very interesting state of charge table for the lifepo4 chemistry.
Will it provide an equally precise indication for 3.2 v batteries if I divide the 6 volt column in 2 ?

yes a 6v LiFePO4 uses 2x cell in series. divide chart by half gives accurate readings.

for my example I'm going to stick with 12v LiFePO4 which is the most commonly adopted form factor. reason why 12v LiFePO4 is taking the motorcycle world by storm .. it's the cheapest way to drop weight on an already light motorcycle. on test mule BMW R80G/S .. by installing a EarthX ETX36 (14AH) .. dropped 20lb vs oem lead acid battery. typical weight loss is under 10lb. R80G/S weight 368lb dry .. dropping 20lb is equal to 1/2 weight of my luggage.

whereas dropping say 50lbs on a 5k lb car is not that big a deal .. so the only market for 12v LiFePO4 for cars is folks like Porsche 911 .. which offered world first OEM LiFePO4 battery at $2,000.

12v LiFePO4 reaches full charge at 14.6v .. after resting overnight will drop to about 14.25v with no BMS .. with cell balancing boards resting voltage will be 13.85v range.

at first current draw .. voltage will quickly fall to 13.3v .. then stay almost flat until near end of discharge. 13.3v (90%) to 12.85v (20%) for max life don't drop below 12.85v or charge above 14.4v.

typical LiFePO4 charge/discharge cycles 2,000+ if above guidelines are followed. ALL PB vehicles charging systems charge 13.8v to 14.2v with minor variations to adjust for hot or cold climates. most moderate motorcycles use a permanent magnet charging system that uses a rectifier/regulator combo. typical charging voltage is 14.4v.

ALL 12v LiFePO4 uses 4 cells in series ... some use 26650 cells in series/parallel configuration. other mfg will use 4x flat prismatic cells to achieve same voltage.

for flashlight use simply divide by four .. all charging parameters are the same.
a single cell LiFePO4 voltage matches with a single primary lithium cell.

LiFePO4 tolerates wild overcharge/abuse .. it takes 29v+ to destroy a 12v LifePO4 battery. yes it's possible to abuse LiFePO4 to point of thermal runaway .. but in practice very difficult to do. more likely battery will simply melt ..

most common reasons for 12v LiFePO4 to melt are:

1. dead short due to poor installation
2. internal strap failure from poor mfg .. this problem was seen on very early LiFePO4
3. rectifier/regulator failure with output at full tilt .. voltage up to 70v delivered to battery
4. installing a too small AH LiFePO4 on a high wattage output alternator which results in charge rates high as 7C. charge rate should not exceed 4C

 

[cy]

I don't know that this drop in voltage necessarily represents a significant loss of capacity. From what I've seen, it seems to be similar to what happens with NiMH if you 'rest' it after charging. You will experience a drop in the open circuit voltage of the cell. But that doesn't mean the cell has suddenly lost significang capacity compared to one that was used without resting. Also remember that, with LiFePO4, this drop from float to plateau voltage occurs much more quickly than it occurs with NiMH. This suggests that any amount of energy that could be lost when a battery rests is negligible.

switching to single cell voltage for this example .. full charge is reached at 3.65v with a resting voltage of 3.56v range .. but at first draw voltage almost immediately drops to 3.325v range .. then discharge curve remains almost completely flat. initial drop means almost nothing in terms of useful AH capacity.

 

[Viking]

Thanks for the explanation.
I will copy that table

 

[StorminMatt]

Another advantage for me anyway, is the very flat discharge curve. For powering heavy current 28 Volt nominal searchlights, I can get only about 40% of the rated capacity out of the pack before Voltage falls below acceptable whereas I can get 80 to 90% capacity out of a pack with LiFeP04 before Voltage falls too low. You can't simply add another cell to the Lipo pack because you would fry the electronics.

I have to admit that a flat discharge curve is something I REALLY like. Admittedly, that's probably a carryover from my use of NiMH. But lately, I've started to use some IMR and ICR batteries in some lights. And I just REALLY dislike the discharge curves of those batteries. Especially noticeable is the way that an IMR 14500 can't run an SC52 on high (500 lumens) if it is more than maybe a third to a half discharged. Simply put, it just can't kick out the voltage at that point. NiMH or LiFePO4 would NEVER do this. The same goes for my Maelstrom MMU-X3, which gets noticeably weak as you approach the end of the discharge cycle. NiMH and LiFePO4 just don't do this. In fact, when I run my 75W incan from four A123 Systems 26650s, it's like I have fresh batteries until they are empty. Now what we REALLY need is a 3.7V Li-Ion that doesn't experience that characteristic voltage drop with decreasing SOC that you get with most Li-Ion.

 

[bob_ninja]

Interesting albeit older discussion. I am surprised I don't find more talk of Fe chemistry.

I have 10+ year old A123 based power tools packs. They still work albeit with clearly higher IR and lower power/capacity. They have proven to be fairly resilient to abuse including not being used for 6+ months, left in -30C temps, etc.

What I find very appealing is their tolerance for abuse. Simply put, most people (wife, kids, etc.) do not monitor cells as closely as we do. Once my wife simply left a light on in basement for hours and forgot about it. I was horrified. Of course cell was dead but at least it didn't burn/explode.

Fe cells ability to handle abuse, both discharge and charge extremes, so well is fantastic and much safer. I don't think that other chemistries are quite as good. You can find R/C forums to see "real" abuse and they found Fe cells behave like a tank.

I am actually puzzled to find so little choice of Fe cells now that A123 is gone. Just checked Tesla and it is using Panasonic NCR 18650A, improved Cobalt chemistry. So I guess Iron took a back seat.

Anyway what is THE safe Li chemistry these days? That is actually available, easy to find.

 

[night.hoodie]

Anyway what is THE safe Li chemistry these days? That is actually available, easy to find.

Efest LiMn? I don't know how safe anything can be with that much energy. Also, not sure how easy or available you need LiFePO4, but I got mine, will be getting more, from fasttech... just takes a little patience for the shipping.

 

[bob_ninja]

Me too, looking at FT, DX, etc.
The trouble is most of them are XxxxFires which I really want to avoid.
FT has cool/world something brand that I am not familiar with.
Managed to find 1 test from HKJ that was positive, so will be trying that brand.
Anyway don't see a lot of choices, that is all.

Especially these days with shipping issues

 

[night.hoodie]

Anyway don't see a lot of choices, that is all.

Especially these days with shipping issues

I don't either, and I think this is odd if China has adopted this chemistry for ordinary consumption. Maybe a national will be kind enough to help ferret out a way to get access to these unknown brands that I'd expect are pretty common there.

I started a list in another thread some time ago, have not found more:

I found a few other LiFePO4 makers... (in no particular order):
1) Coolworld
2) Coolook
3) Soshine
4) Imren
5) Powerlion