LiFePo - LiPo protection and such

[Moriambar]

Hi.
I am accustomed to the fact that LiPo batteries require special care and attention, especially regarding overcharge and overdischarge protection, but also overcurrent; I also know that there are specialized circuits and that many LiPo batteries contain such protections inside the battery case itself.
The extra care applies also when recharging a battery (e.g. from a power supply).
But what about LiFePo? I know they are safer (but their energy density is lower), but what does this mean? Can I use them as "normal AA" (ofc keeping in mind the different voltage and size), or should I build/buy specific protection circuits?
Are they chargeable the same way as the lipo (except adjusting the specific parameters as per the datasheet)?

Cheers

[Siwastaja]

The same applies; you need to keep the voltage between the allowed limits, preventing overdischarge and overcharge.

LFP being safer is a myth. At cathode chemistry level, it is true with significantly higher thermal runaway onset temperature and significantly lower released energy in such event, but at product level, there are more factors involved, and most LFP cells on the market are made by substandard companies.

[Moriambar]

The same applies; you need to keep the voltage between the allowed limits, preventing overdischarge and overcharge.

LFP being safer is a myth. At cathode chemistry level, it is true, but at product level, there are more factors involved, and most LFP cells on the market are made by substandard companies.

Ah ok, that's interesting.
I won't buy them then: having the protection circuit integrated in the battery, plus the better energy density basically make me wonder if there is any reason, after all, to buy LFP.

[Jwillis]

All lithium ion batteries don't charge well in series unlike other rechargeable batteries such as NiMH and lead acid . This is where balancing the charge with a controller is required. Essentially what the controller does is monitor all the batteries that are in series and charges each one as an individual. This prevents over charging of the leading batteries over the trailing batteries . These controllers also measure the temperature and how well each battery in series is charging. 
 In a parallel configuration theirs not much problem as long as the charge rate  doesn't exceed what the battery can except over a period of time .   Because all batteries have an inherent resistance and  the charge rate is to high it will produce excessive heat. If enough heat is produced it doesn't matter what type of battery it is the excessive heat can shorten the life of the battery or cause a catastrophic failure .
So its best to charge  LiFePo as you would any other rechargeable  lithium battery. Especially in a series configuration.

[Moriambar]

All lithium ion batteries don't charge well in series unlike other rechargeable batteries such as NiMH and lead acid . This is where balancing the charge with a controller is required. Essentially what the controller does is monitor all the batteries that are in series and charges each one as an individual. This prevents over charging of the leading batteries over the trailing batteries . These controllers also measure the temperature and how well each battery in series is charging. 
 In a parallel configuration theirs not much problem as long as the charge rate  doesn't exceed what the battery can except over a period of time .   Because all batteries have an inherent resistance and  the charge rate is to high it will produce excessive heat. If enough heat is produced it doesn't matter what type of battery it is the excessive heat can shorten the life of the battery or cause a catastrophic failure .
So its best to charge  LiFePo as you would any other rechargeable  lithium battery. Especially in a series configuration.

Thank you. I currently am considering a single-cell configuration but knowing this is incredibly useful

[Siwastaja]

Li-ion cells charge in series very well because self-discharge currents are very low (typ. 3%/year) and coulombic efficiency extremely high (typ. 99.95%), and offsets of these parameters between cells even lower. But, because at the high state-of-charge side, they can't properly self-balance acting like zener diodes, like some other chemistries can, you need external monitoring to prevent overcharging. And, because overdischarge resulting in copper dissolution is actually dangerous as well, you need external monitoring to prevent overdischarging as well.

The funny thing is, in a proper li-ion pack, such monitoring and balancing does very little work! Hence people and even reputable companies such as Robert Bosch have got away with "dumb" series strings with no balancing, no cell-level monitoring. Nowadays, cell-level UV/OV protection is almost always used and highly recommendable, though, despite it looking wasteful because it's sitting there doing mostly nothing.

This all applies to LFP and other li-ion cathode chemistries in use (mostly NCA and NMC these days, it seems).

[Siwastaja]

Thank you. I currently am considering a single-cell configuration but knowing this is incredibly useful

If at all possible, use single-cell configuration, even if that requires a boost converter to get 5V, or similar.

Go for multicell if you need significantly higher unregulated voltages (think about motor controller running at raw 20-30V range), for example.

[sleemanj]

LiFePO4 is much much less likely to turn into a fireball.  That doesn't mean you can mistreat them, if you do they will be irrevocably damaged, but, they will at least not make your house burn down.

Here's a video showing some cells of various types being mistreated.

[sandalcandal]

Plenty of advertising videos show LFP smoking and venting while comparing them to fireballing NMC or NCO LCO cells but they're a bit misleading. The intensity of thermal runaway in LFP is lower so self ignition is less likely but the vented gasses are still very flammable (and toxic). Other people have mentioned it already but don't get complacent with LFP.

Unfortunately not an open access paper but a recent (Jan 2020) academic study here: https://www.researchgate.net/publication/334664231_A_comprehensive_investigation_on_the_thermal_and_toxic_hazards_of_large_format_lithium-ion_batteries_with_LiFePO4_cathode

Edit: Open access paper from same above authors but focused on fire behaviour https://www.researchgate.net/publication/334575740_A_new_exploration_of_the_fire_behaviors_of_large_format_lithium_ion_battery

[Siwastaja]

Also comparing a low-quality no-name Chinese LFP brick to a low-quality no-name Chinese LIPO pouch is a misleading comparison because with NCA/NMC, you do have a choice of buying Samsung, LG or similar - and even at lower cost per kWh than the LFP bricks!

Set a price point, pick highest quality products within that price, then compare safety. Then it would be some Thundersky vs. Samsung, not Thundersky vs. Hu Long Whatever Co. Ltd lipo cell.

All my attempts trying to get a Sony or Samsung li-ion cells to ignite through electrical abuse have failed, this includes applying 10A charging current at some 30 volts to a 18650 cell for some hour or so!

And this is also what I base my claims on, I remember a research paper but don't remember the authors or title offhand showing COTS LFP cells venting significantly more aggressively than COTS LCO cells during same electrical abuse conditions. The reason was, LFP cells didn't have passive cell-level protection, LCO cells do have because they have to due to the more dangerous cathode chemistry.

[Moriambar]

Also comparing a low-quality no-name Chinese LFP brick to a low-quality no-name Chinese LIPO pouch is a misleading comparison because with NCA/NMC, you do have a choice of buying Samsung, LG or similar - and even at lower cost per kWh than the LFP bricks!

Set a price point, pick highest quality products within that price, then compare safety. Then it would be some Thundersky vs. Samsung, not Thundersky vs. Hu Long Whatever Co. Ltd lipo cell.

All my attempts trying to get a Sony or Samsung li-ion cells to ignite through electrical abuse have failed, this includes applying 10A charging current at some 30 volts to a 18650 cell for some hour or so!

And this is also what I base my claims on, I remember a research paper but don't remember the authors or title offhand showing COTS LFP cells venting significantly more aggressively than COTS LCO cells during same electrical abuse conditions. The reason was, LFP cells didn't have passive cell-level protection, LCO cells do have because they have to due to the more dangerous cathode chemistry.

I'm sorry but I don't understand COTS, LCO, NCA, NMC and actually I am confused.
Since you seem to know your stuff, let me rephrase my question: why should anyone use LiFePo batteries instead of LiPo batteries, giving the cost and (apparently) the same need of a similar protection circuit (that many LiPo have integrated in the case anyway)? Is it just for the reduced flammability? I mean comparing, eg, the same high quality brand in LFP and LiPo

[Siwastaja]

I'm sorry but I don't understand COTS, LCO, NCA, NMC and actually I am confused.
Since you seem to know your stuff, let me rephrase my question: why should anyone use LiFePo batteries instead of LiPo batteries, giving the cost and (apparently) the same need of a similar protection circuit (that many LiPo have integrated in the case anyway)? Is it just for the reduced flammability? I mean comparing, eg, the same high quality brand in LFP and LiPo

Sorry, I try to avoid abbreviations but sometimes use them anyway.
COTS = commercial off-the-shelf.
LFP, LCO, NCA, NMC = lithium iron phosphate, lithium cobalt oxide, nickel-cobalt-aluminium, nickel-manganese-cobalt, different li-ion cathode chemistries. Anode is always carbon in some form (graphite, for example), recently also silicon being added seems to finally work out.

Do note, "lipo" does not exist. It's all li-ion. We professionally call different form factors: prismatic, cylindrical, or pouch. Since a really different thing called "lipo", with completely solid electrolyte, failed to emerge in 1990's, but the name was too good to be forgotten, the term was later picked up for reuse by marketing teams just to describe the pouch form factor.

So "lipo vs. lifepo" is just completely nonsensical. LiFePO4 (I like to say LFP instead) can be prismatic, cylindrical, or pouch as well, it's just the form factor.

Your question, why would anyone use LFP, is a good one. It has potential for better safety (which needs to be realized by not wasting it by lacking in safety otherwise); it also uses no cobalt which is an expensive conflict mineral. It seems to me, almost no one is using LFP, the energy density is just so poor and this also translates to cost, even if the materials are cheaper, you need more of them. It seemed like a good idea in early 2000's but picked not much commercial interest outside China, and even within China, it's mostly used by fairly small players. Also, the amount of cobalt used by "traditional" li-ion went down significantly when NCA and NMC superseded the classic LCO chemistry during the last decade or so.

[sandalcandal]

Some small additions to what Siwastaja has said:

CoTS=commercial-off-the-shelf, as in a ready made component or sub-system (or even complete system) as opposed to one that was custom made to order (in-house or outsourced).

Chinese manufacturers and academics sometimes seem to refer to nickel-manganese-cobalt using the abbreviation NCM rather than NMC for some reason.

LFP has potential to be cheaper (and more ethical) due to lack of cobalt and nickel in the chemistry, particularly in terms of total lifetime throughput (cycle life) in less demanding applications. CATL is a "big" player that does significant LFP production but mostly for EVs and ESS (energy storage systems) not consumer gadgets. There is a lower cost version of the Tesla Model 3 made in Shanghai that uses LFP produced by CATL. Haven't checked but the $/kWh (cost-capacity efficiency) for that Model 3 is likely higher as Siwastaja mentioned despite lower up front costs. All the LFP I've seen are worse than modern NMC and NCA in terms of cost for capacity.

Edit: LFP also has the benefit of being able to be configured to roughly match common 12V, 24V, 48V lead acid battery voltage ranges due to the 2.0V to 3.60V range per cell compared to 3.0V(2.5V) to 4.2V per cell in other higher energy li-ion chemistries. So you'll often see drop-in replacements for lead acid battery systems using LFP.

[Moriambar]

I'm sorry but I don't understand COTS, LCO, NCA, NMC and actually I am confused.
Since you seem to know your stuff, let me rephrase my question: why should anyone use LiFePo batteries instead of LiPo batteries, giving the cost and (apparently) the same need of a similar protection circuit (that many LiPo have integrated in the case anyway)? Is it just for the reduced flammability? I mean comparing, eg, the same high quality brand in LFP and LiPo

Sorry, I try to avoid abbreviations but sometimes use them anyway.
COTS = commercial off-the-shelf.
LFP, LCO, NCA, NMC = lithium iron phosphate, lithium cobalt oxide, nickel-cobalt-aluminium, nickel-manganese-cobalt, different li-ion cathode chemistries. Anode is always carbon in some form (graphite, for example), recently also silicon being added seems to finally work out.

Do note, "lipo" does not exist. It's all li-ion. We professionally call different form factors: prismatic, cylindrical, or pouch. Since a really different thing called "lipo", with completely solid electrolyte, failed to emerge in 1990's, but the name was too good to be forgotten, the term was later picked up for reuse by marketing teams just to describe the pouch form factor.

So "lipo vs. lifepo" is just completely nonsensical. LiFePO4 (I like to say LFP instead) can be prismatic, cylindrical, or pouch as well, it's just the form factor.

Your question, why would anyone use LFP, is a good one. It has potential for better safety (which needs to be realized by not wasting it by lacking in safety otherwise); it also uses no cobalt which is an expensive conflict mineral. It seems to me, almost no one is using LFP, the energy density is just so poor and this also translates to cost, even if the materials are cheaper, you need more of them. It seemed like a good idea in early 2000's but picked not much commercial interest outside China, and even within China, it's mostly used by fairly small players. Also, the amount of cobalt used by "traditional" li-ion went down significantly when NCA and NMC superseded the classic LCO chemistry during the last decade or so.

wow, thanks.
I really wish there were an EEVBlog (or some other video) explaining and experimenting in detail with all of this, especially NMC and NCA which I never heard of

[Siwastaja]

Just remember that a complete cell is much more than just cathode, yet we tend to talk about cathode chemistries only.

The cell is cathode material coated on copper foil, anode material on aluminium foil, separator material inbetween (even this is non-trivial, quite high-tech actually, much more than just a perforated plastic film!), electrolyte (which still sucks, there is still no replacement for the lithium salts in highly flammable organic solvent mixture), and finally, packaging. All aspects of the cell affect the final product.

The problem in all cathode chemistries mentioned is they contain oxygen, and hence can go on and release heat without external source, such oxidizing fires are hard to put out because removing the source of external oxygen does not help. The oxygen in LiFePO4 is most stable of the bunch mentioned; the oxygen in LCO is least stable. Other cathode chemistries fall between these. (I'm ignoring LTO for now because the energy density sucks even worse than LFP.)

Note, LCO, LTO and LMO have the Oxide in their name, NCA and NCM are oxides as well, they "should" be called LNCAO and LNCMO to be pedantic and logical but somehow we engineers love three-letter acronyms.

LCO is the chemistry where it all started in late 1980's/early 1990's, now it's becoming quite deprecated because NCA/NMC seem to be somewhat better in almost all regards.