LifePO4 vs 18650 Lithium

[jubjub2659]

Hello!

I am just wondering what peoples opinions are with regards to comparing lifePO4 cells and the 18650 lithium cells.

From what I can tell the lifePO4 cells seem like the better choice as they have more cycles and have a higher nominal voltage than the 18650. Yet from what I see everyone is talking about the 18650 cells specifically in the 52v package. Even with the 52 volt set up the max voltage of the the 52v 18650 is almost exactly the same as a 48v lifePO4 pack. The only downside I see to lifePO4 is they cost a little more but with over double the life cycles I don't really think that is a big deal.

I am pretty new to all of this so I just wanted to see what other folks think!

Thanks!

 

[footloose]

I've used batteries made up of A123 26650 LiFePO4 cells and several types of Samsung and Panasonic 18650 cells (NMC and NCA chemistries). All are good, but there are trade offs involved in making choices among them.

Probably nothing on ES incites stronger opinions than battery chemistry, but these graphics kind of sum it up for me.
LiFePO4 is terrific, but it has lower "specific energy"... which means you need to carry more poundage of battery for equivalent energy storage. It has a longer life span, but batteries are evolving so quickly I kind of expect anything I buy today I'll probably want to replace with next generation in a year or so, so life span not so crucial.

Right now, for my electric bicycle purposes, the 18650 Lithium Ion cells seem the better choice. Good balance of cost / energy / life span etc.



source: http://batteryuniversity.com/learn/article/types_of_lithium_ion

 

[Sunder]

18650 is a format.
LiFePo4 is a chemistry.

It's a bit like asking "Do you prefer petrol cars, or hatchbacks?"

In fact, you can get an 18650 LiFePo4 cell:

https://www.aliexpress.com/item/4pcs-Lifepo4-18650-Batteries-3-3V-1100mAh-Rechargeable-Battery-System-For-EV-Tool-Ebike-Original-Free/32502991082.html?spm=2114.01010208.3.1.tM5qYR&ws_ab_test=searchweb0_0,searchweb201602_1_116_10065_117_10068_114_115_113_10084_10083_10080_10082_10081_10060_10061_10062_10056_10055_10037_10054_10059_10032_10099_10078_10079_10077_10073_10100_10096_10070_423_10052_10050_424_10051,searchweb201603_2&btsid=73fb3b7a-ddaa-4e8a-99f8-22a354660f35

You're probably thinking Lithium Ion 18650s, even then, you have a few variations.

Most people when choosing 18650s, want the highest energy density, and prioritise this over power density or cell life.

If you're space tight or weight conscious, and want the longest range, then 18650s is probably your best choice. If you're after a very long life, and don't mind the space/weight issue, then LiFePo4 can be very good value. It will actually lose power density before it loses energy density (I.e. after 3 years, you'll still have near rated capacity - but the voltage will sag under load), so you can preempt this by oversizing your battery, or adding another parallel set when the sagging gets too bad.

 

[dilkes]

From what I can tell the lifePO4 cells seem like the better choice as they have more cycles and have a higher nominal voltage than the 18650. Yet from what I see everyone is talking about the 18650 cells specifically in the 52v package. Even with the 52 volt set up the max voltage of the the 52v 18650 is almost exactly the same as a 48v lifePO4 pack. The only downside I see to lifePO4 is they cost a little more but with over double the life cycles I don't really think that is a big deal.

Individual LiFePO4 cells have a lower voltage than Lithium Ion cells. A "48 volt" LiFePO4 pack is typically a 16s configuration and hot off the charger it's about 57.6 volts (3.6 volts/cell). The closest equivalent for Lithium Ion is the "52 volt" pack which is a 14s config. Hot off charger is about 58.8 volts (4.2 volts/cell). Therefore a comparable LiFePO4 pack would need 14% more cells (and weight).

 

[redilast]

I don't think LiFePo4 provides more cycles. If you decrease the DOD of a non-LiFePo4 cell, you can match the cycle life of LiFePo4 but still have better power density. There is a reason most electric car companies don't use LiFePo4....

 

[dogman dan]

My own personal experience is you might get one more year from lifepo4, than other lithium chemistries.

VERY FEW ride enough to run out of cycles. Really, packs time out because it's inconvenient to store them discharged, cuz you want to ride it anytime. So for most, its charge it, ride, charge it again. And then in about 3 years, sitting in a hot garage some, sitting charged a lot, they time out.

But here is the deal,, the current crop of 18650 format cells are typically a chemistry ( lithium manganese something or other) that works fine, is at least as good for discharge rate as lifepo4, has a HIGHER voltage per cell than lifepo4, and is a TON smaller. This results in very similar li-ion (technically they are both li-ion) packs being a LOT smaller and lighter than lifepo4, which needs more cells to reach 36v.

But this,, I'm talking about a bike or scooter battery. The size and weigh matter a lot, and it's kind of hard to use it so gently with either chemistry. If you want to run your house on it, I'd still choose lifepo4, and a use profile that could potentially get 10 years from the cells.

 

[cycleops612]

18650 is a format.
LiFePo4 is a chemistry.

It's a bit like asking "Do you prefer petrol cars, or hatchbacks?"

In fact, you can get an 18650 LiFePo4 cell:

https://www.aliexpress.com/item/4pcs-Lifepo4-18650-Batteries-3-3V-1100mAh-Rechargeable-Battery-System-For-EV-Tool-Ebike-Original-Free/32502991082.html?spm=2114.01010208.3.1.tM5qYR&ws_ab_test=searchweb0_0,searchweb201602_1_116_10065_117_10068_114_115_113_10084_10083_10080_10082_10081_10060_10061_10062_10056_10055_10037_10054_10059_10032_10099_10078_10079_10077_10073_10100_10096_10070_423_10052_10050_424_10051,searchweb201603_2&btsid=73fb3b7a-ddaa-4e8a-99f8-22a354660f35

You're probably thinking Lithium Ion 18650s, even then, you have a few variations.

Most people when choosing 18650s, want the highest energy density, and prioritise this over power density or cell life.

If you're space tight or weight conscious, and want the longest range, then 18650s is probably your best choice. If you're after a very long life, and don't mind the space/weight issue, then LiFePo4 can be very good value. It will actually lose power density before it loses energy density (I.e. after 3 years, you'll still have near rated capacity - but the voltage will sag under load), so you can preempt this by oversizing your battery, or adding another parallel set when the sagging gets too bad.

+1

yes, one is a cannister pack dimensions, the other a compound~

anyhoo,

as some here say, lifepo4 is 5x better than lead acid of yore, and suits a similar mindset. faint but solid praise. Given the dominant chinese domestic market is still big on lead acid, lifepo4 seems an odds on replacement choice for this huge, mostly utility bike, market.

A neglected matter is complexity/cell capacity size. Even a 1kwh lifepo4 (lfp) pouch pak is a mere 12 cells (or 24 if u wanna boost/double~ c-rates).

Try that with with 18650. if optimistically, 3000mah at 3.7v nominal?, then its about a hundred cells of 11wh each. Its very messy, and that usually means cost and hassle. u need to be nerdy to mess with them.

I have yet to hear how a bms balances properly with so many cells?

So it seems odd lipo pouches are little discussed? - partly, the best of both worlds for limn 18650 fans.

I hear they are basically the same chemistry as limn 18650s, in pouch form - any cell size u want and simpler manufacture and pak assemblies.

https://www.google.com.au/search?q=3.7v+20ah+pouch&espv=2&biw=1408&bih=682&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwixu43kzLzQAhUJI5QKHftqDNkQsAQIGQ&dpr=1.25

I only have a small problem with 18650/limn. The numbers they claim are utter bullshit. Maybe they are still better, but i hate feeling lied to.

For starters, the bottom 20% of claimed capacity is unusable, yet it seems fine to simply ignore this?

There is also the small matter that a 10ah 36v 1C limn is presumably 360w for an hour, right? Not only is it nothing of the kind, as above. After a few miles, there is no way it will produce the full 360w.

Silly me I am told. One simply gets a 15ah limn, and it will do that fine, just like a 10ah lifepo4. OKAAAY?

 

[999zip999]

Quality. Most lifepo4 cells are 2c. Some 20c. Today the way things are going a high quality 18650 pack with a high-quality cell like Samsung 25r cell or ?Something over rated so it will last. If the battery pack maker can't tell the cell don't buy. I don't mean a Tesla like cell. So printing or a write up b******* I mean what cell.
What are you using it for ? Voltage ,controller ,amps, motor, terrain and weight. Price ?

 

[DVDRW]

Some numbers

1kw/h 18650 weights 4.8kg
1kw/h lifepo4 weighs 10.15kg

1kw/h 18650 volume is 1.6 Litres
1kw/h Lifepo4 volume is 5.52 Litres

1kw/h 18650 cost is 250€
1kw/h Lifepo4 cost is 526€

Calculation based on:
Lifepo4 https://www.ev-power.eu/LiFePO4-small-cells/LiFePO4-High-Power-Cell-3-2V-20Ah-Alu-case-CE.html?cur=1
18650 Liion https://eu.nkon.nl/rechargeable/18650-size/panasonic-ncr18650pf-3-7v-2900mah.html
With shipping to Estonia.

Lifepo4 is 3x more volume, 2x heavier and 2x more expensive.

No thanks

Lifepo4 cycle life is ~700 cycles @ 2C (been there done that)
To use Lifepo4 cycle life benefit @ lower C rate, battery must be very big and heavy.

Here someone tested 18650 25R cell cycle life got more than 1800 cycles.

 

[cycleops612]

I am not going to argue with people who dont even address the argument.
An actual product list:

http://www.bestgopower.com/images/products/BP-list.jpg
(for simplicity, suggest compare 10ah pak in each chemistry).

nothing like the claims above, which are meaningless even if true if the ratings are BS.

Yet again, that 20% of rated capacity is utter bs id regarded as a trifle not worthy of mention or response.

What consumers buy is called marketing. 18650 limn are a marketers dream. They sound great on paper. Its easier and cheaper to create a nominal 36v (say) pak, which it is, for a few minutes after a full charge. Ask any tessla driver about "sag". Its not my idea of a car. Horses get tired, cars dont. We have moved on.

For hybrids which i suspect constantly work within the 20% & 80% charge levels using the ice dextrously, yep, good choice.

I repeat. They may be ok, but i hate lies.

 

[Hillhater]

You seem to ignor selective facts..?
For any given capacity, the LiFePo4 pack will be bigger, heavier, and more expensive....all pretty relavent on an Ebike !
Voltage sag is a function of internal resistance (directly related to C rate) for all cell chemistrys .....
....so, unless you invest in the premium high rate LiFePo4 cells ( at an premium cost) then for packs of similar capacity and C rate you will get similar voltage sag effects.
If you plan on using every last Whr in the pack , and worry about that last 20% in the 18650s, then just add more capacity to the pack....it will still be smaller , lighter, and cheaper than the LiFePo4 pack.

 

[cycleops612]

FYI folks
some good stuff here:

https://books.google.com.au/books?id=Mbd7BgAAQBAJ&pg=PA84&lpg=PA84&dq=biggest+volume+18650+cell&source=bl&ots=ZH4sO2tRss&sig=5wc1bf35nGFuZz68yawjRRmbiZk&hl=en&sa=X&ved=0ahUKEwjntu3l9sPQAhWGqJQKHcmvDTYQ6AEIVDAJ#v=onepage&q=biggest%20volume%2018650%20cell&f=false

 

[dogman dan]

Yes, sag will be the same if C rate is the same.

The problem though, is the lies in the c rate, and sometimes capacity too.

So it gets difficult to figure out without individual testing by somebody. 3c lifpo4 may sag less than 3c li-ion because one vendor or the other is the bigger liar.

Ok, maybe not outright lies,, but fact is, what you are dealing with is the sag of a completed pack, while that spec is a lab test of one cell. That' complete pack is including the resistance of the nickel straps, the spot welds, the bms, some crummy soldering, and even the shitty plug on the end it came with.

So even same cells, it can be differing results out on the road.

But some generalizations are true. Both made of 18650's, a lifepo4 pack is larger in volume and weight than the li-on 18650's. Part of that is very simple, 13s makes a 48v pack in one, while lifepo4 needs 15s for the same voltage. so at the very least, just the same voltage is two more rows of cells for lifepo4. Then there may or may not be more capacity in the same size li-ion cell. But with good cells, the li-on 18650 pack will need one or more less cells in each of the 13 rows. So its gotta be smaller and lighter. Its at least 8 cells less, or more in a 4p pack. If one is 4p and the other 5p, its 21 cells smaller.

 

[Punx0r]

FYI folks
some good stuff here:

https://books.google.com.au/books?id=Mbd7BgAAQBAJ&pg=PA84&lpg=PA84&dq=biggest+volume+18650+cell&source=bl&ots=ZH4sO2tRss&sig=5wc1bf35nGFuZz68yawjRRmbiZk&hl=en&sa=X&ved=0ahUKEwjntu3l9sPQAhWGqJQKHcmvDTYQ6AEIVDAJ#v=onepage&q=biggest%20volume%2018650%20cell&f=false

It's not a bad book, but the assertion that 18650 packs are inherently less reliable than those made of large-format pouches is demonstrably false. It's based on the simple relationship between part count and likelihood of failure of the assembly, but this is an over-simplification and. The large number of connections in does make the occurrence of a faulty connection more likely (assuming the reliability of the process for making 18650 connections is identical to that for pouches, which it's probably not), but the 18650 pack is much more tolerant of a single faulty cell connection - it will continue to have a near-design life whereas the large pouch pack would immediately require replacement/repair.

It also negates the different probabilities of failure of the different components. If cell failure is far more likely than connection failure, then it matters much less how many connections the pack has. 18650 cells are more reliable and, as above, the pack can tolerate the odd failure cell. The reliability of 18650s seems in part to be how well developed and mature the manufacturing process is, and also inherently because of their small size - it's easier to make a small thing defect-free and easier to detect small defects (ref: microchip wafer manufacture).

 

[dogman dan]

Failure rate of a cell has little to do with the type of can its living in. round can, square can, naked pouch, all can be manufactured poorly. A contaminated cell fails every time.

The most developed and perfected manufacturing processes at this moment could be the 18650 can.

 

[Punx0r]

Yeah, I was assuming good 18650's - the sort an OEM buys from a tier1 supplier, not an ebiker buying from alibaba

 

[cycleops612]

Jeez guys, I cant cut and paste, but r u sure u read the bit on e.g.7000 cells for tessla & the zillion welds that needs?

For goodness sakes. Its a no brainer that the slightest glitch can make it uneconomic to repair and more likely to fail. A bookie could tell u that.

As always, not one response over the years, on how on earth BMS is supposed to individually balance cells? My guess is its an inconvenient truth.

 

[Punx0r]

Yes, 7000 cells & 14,000 welded connections. So? Try reading my post above...

The proof of the pudding is in the eating: Tesla offer by far the longest warranty on its EV packs. Do you hear user reports of battery failure being common?

Here's the ascending order of how you assess if something is designed well for a given application:

Speculation (here's where you are)
Sound theory
Simulation/modelling
Physical testing
Field proven (here's where Tesla is)

A BMS does not balance individual cells, only cell groups. Nor does it need to. This is the same whether you have 3P large pouch cells or 74P 18650 cells.

What are you trying to prove and why?

 

[cycleops612]

Most of the grief we see here is folks forgetting its just a bicycle.

500w, mid drive and under 120kg gvm, is within spec, just.

Riding using 360w is plenty to stay under wind barrier speed and tackle any hill at a run at least in the right gear. 140w in reserve is nice if needed.

this sounds a fine balance to me - about an hour of full 360w power, lightish, longevity:

http://www.pingbattery.com/servlet/the-8/lifepo4-lithium-phosphate-iron/Detail

Specifications:

Suitable Wattage of Motor: up to 500 Watt
Suitable Motor Controller: up to 25 Amps
...
Voltage: 36 Volts (38.4 Volts Nominal)
Capacity: 10 Amp Hours
Dimension: 135*105*155 mm / 5.3*4.1*6 inches
Weight: 3.7 kg / 8.2 lbs
Charging Voltage: 44-45 Volts
Charging Current: <8 Amps (<10 Amps with High Rate BMS)
Rated Discharging Amperage: 10 Amps
Max Continuous Discharging Amperage: 20 Amps (50 Amps with High Rate BMS)
Maximum Discharging Current: 60 Amps
Discharging Cut-off Protection: 25-30 Amps (65-70 Amps with High Rate BMS)
Lifecycle of the whole pack: >85% capacity after 1000 cycles. Lifecycle of single cell: >85% capacity after 1500 cycles, >70% capacity after 3000 cycles. (<1C discharge rate and <1C charge rate), Averagely 2000 Cycles

Ping do not BS. They value their reputation too much, hence their prices. Their customers want no BS mainly. Time is money.

An 18650 10ah wouldnt have a hope. Its 20% less capacity for starters, as nobody refutes.

c-rate perceptions are biased by the fact that ALL 18650 limn paks have parallel cells (which seems to improve c-rates, or it does with lifepo4 pouches), whereas lfp pouches rarely do. Ping and others use limited paralleling to achieve very decent c-rates (eg 6C max above using 2p). If C-rates are your thing, LFP can help as well as limn, but...

its a balance,

ebikes need range too, and no chemistry likes doing sustained high c-rates. RC lipo guys replace every 6-12 months i hear. I want a strong v8 battery that does its job reliably for many years, not a prima donna F1 battery. If that means an extra kilo, so be it. Its arguably the most important component. Steel forks are heavier too, but many here use them.

As voltage droops on limn, so c-rates must increase to maintain power during a journey. Not a good look for the battery OR the circuitry (more current).

I am not alone in saying, that to equal the above lfp, a limn needs to be 15ah (I make it 16ah), all pros and cons bar longevity considered. If u compare the weight of a 15ah limn w/ a 10ah lfp, i doubt there is much weight/volume advantage to limn cannister paks.

 

[fechter]

I have a Li-ion pack and an A123 pack that are roughly the same physical size. The Li-ion pack has about 2x the capacity and is significantly lighter than the LiFePO4 pack. A123 claims higher cycle life and can dish out way more amps without heating.

 

[dogman dan]

Well, lets not get carried away here, calling a quality assembled 18650 pack the same chance to fail as some $400 pack off alibabba.

I'll take 1400 perfect robo spot welds, over a few hundred dodgy hand done ones.

Individual cells in a parallel group do not balance individually. So one bad one in the group can be a disaster. Again, back to QC of the cells. For $400 you are not getting a pack with perfect welds, and cells with a low chance of failure.

Re the ping battery. Buy the 15 ah, and preferably the 20. I have about 10,000 bike miles on pings, and loved em. But I did not push them so hard, like the 10 ah will be on a 20 amps controller. They do perform really well though, when a 15 ah is asked to cruise 20 mph.

 

[Punx0r]

An 18650 10ah wouldnt have a hope. Its 20% less capacity for starters, as nobody refutes.

The only person I can recall claiming this is you...

IIRC Ping batteries are good for 2C max and anything more equals a very early death. Run one at 6C if you wish...

Don't compare PC lipo to quality 18650 li-ion can cells. Not sure why you keep say LiMn, 18650's using several different chemistries, including LiCo, same as RC "lipo".

Nothing wrong with a *quality* LiFePo4 pack in the right application. It will last years and do lots of cycles if well made. It will just be twice as big and heavy.

 

[cycleops612]

An 18650 10ah wouldnt have a hope. Its 20% less capacity for starters, as nobody refutes.

The only person I can recall claiming this is you...

IIRC Ping batteries are good for 2C max and anything more equals a very early death. Run one at 6C if you wish...

Don't compare PC lipo to quality 18650 li-ion can cells. Not sure why you keep say LiMn, 18650's using several different chemistries, including LiCo, same as RC "lipo".

Nothing wrong with a *quality* LiFePo4 pack in the right application. It will last years and do lots of cycles if well made. It will just be twice as big and heavy.

re the 20% thing, what a coincidence! As I stated, u r the first to say it isnt. I await confirmation by others. I did a sniff around, but didnt find any timely info.

I am happy to be corrected, but its sure taken years to get a response on it.

my understanding is the normal bms lvc protective cutoff point equates to an 80% discharge level.

In the example i gave of a 1c 10ah limn pak, its pretty useless on a 350w bike after 10 minutes of 350w draw, as it is on my panasonic cell 12AH rig - voltage droop.Its not news lifepo4 has big edge in its flat voltage discharge curve.

The lvc simply formalises matters by forbidding u from even trying to use it.

re the 6c thing, pretty true of any batteries claimed C rating. They need a huge grain of salt. But i think the moral is they can in bursts if u must, but only short bursts, which still has its uses. Maximising cycles isnt everything. Call it an emergency power reserve.

For longevity, your battery should be sufficient for 1c discharge to power normal use for a normal period.

as i said, ping are reputable, so their claims are as credible as it gets.

 

[cycleops612]

Just saying, my impressions. A reasonable common sense rule of logical thumb is the pack makers are liable~ for warranty/bad rap, so their default BMS settings are a good benchmark of whats ok. Its their ass if there is damage by the user.

lfp pouch individual CELLS usually claim 3c. Pack assemblers usually limit to 1.5c max via bms. Parallel cell pouch makers (ping etc) usually ~double this to 3c. It seems not to matter if its 2p or 3p - the bms settings seem the same. They will usually state the effect of different c-rates on longevity. Fair enough.

It bears noting, the controller can over-ride bms settings at times. It may have a draw limit of the equivalent of 1.2c from your battery e.g.

Similarly, the controller may refuse to draw current when voltage drops too far.

The higher the claimed c-rate, the more fanciful they are i hear. 30c lipo is nothing like it in reality e.g.

 

[Punx0r]

Sounds like you had a junky LiMn pouch cell pack and because of that are now rubbishing all 18650 chemistries for some reason?

Your understanding of the various components of a ebike electrical system seem to be incomplete and confused...