Is LiFePO4 worth investing in?

I am going to ask very generally in the following context-

I am in a hobby where the limitation of batteries becomes more of weight than cost or size.

Cost would then be the next limiting factor but that can be extended far past weight.

Anyways, while having done some research here and there about LiFePO4 in terms of testing and the typical usability of cells whether it's cylindrical or prismatic, one thing tends to come up from every topic..

That is they keep failing or they do not last that long. Not the packs, but certain cells.

From where i come from, i drive cars with flooded battery that last 5-8yrs before it needs replacing.

In my current hobby, using AGM batteries that go through multiple owners lasting up to 10yrs with rarely any issues.

Then i go to research LiFePO4 and i can say Lithium by default to the public is a customizable hobby so human error can play a role in the outcome of packs.
But.. it seems that practically every custom pack has issues?

That doesn't sound good for lithium technology to me at all..
However, in my situation, i have an advantage if i am correct on this-

Most users who use lithium need it to longevity until the packs die. eBikes, EVs, drones, etc.

In my situation, i fully control the discharge value and time, So, i am "assuming" LiFePO4 would be great at controlling it's durational life rating.

I have products which will shut off loads 100% to prevent cells from being discharged too low.
Over-charging would be impossible as they would be auto-charged via alternator set lower than required voltage (and manually charged later).

Balancing would be done when manually charging.

So, in my situation, while i see many issues with lithium such as poor QC, few dead cells here n there, lower than rated specs, etc..
As long as i can control what i say above (because my packs are an auxiliary power source, not primary. When in use, they are constantly being actively monitored and protected) then i should continue continue to look into using LiFePO4, correct?

I will not be using any BMS or PCB protection because i can fully control and stop discharge and prevent overcharge from occurring. Active monitors are set in place when using much smaller loads and would be balanced manually.

LiFePO4 cells will last considerably longer than lead as long as they are managed correctly. Not overcharged, not over discharged being essential.

That being said, while LiFePO4 are attractive for weight reduction when compared to lead, they are very heavy/bulky when compared to modern 18650 or large prismatic lithium polymer cells.

What are your requirements for energy/power/size/weight?

Concerning the starter battery in your car (lasting 8 years)...you only drain it for a few seconds to start the car, and then it is immediately topped off. I understand how frustrating it is for so many ebike builders who want a good price on the battery pack (often the most expensive part of a build), but they also want it to last a decent amount of time.

All I can say is select a pack with quality name brand cells, charge at a low enough rate that it doesn't get hot, run it on your bike so that it doesn't get hot, size it so you don't have to drain it all they way down before charging, and only charge to 4.1V per cell.

If you like to draw 25A from a 6P pack, that is about 4A per cell. If you spec a cell that is rated to be capable of 8A per cell, then under your user-profile, the pack shouldn't get hot when you are riding hard.

I am easy to deal with in terms of limitations.

Weight- no limitation.
Being used over 10 AGM batteries as typical requirement.. lithium anything is a miracle technology in that sense.

It would be some AGM and the rest an unlimited amount of lithium cells in terms of what is required to run a load.

Sizing- no limitation.
Again, there is plenty of room even for 1000aH winston cells if i were rich, heh.. but seriously, space is not an issue either.

Power/energy-

I have 2 options so i will attack the harder option first to see where i stand.

I need to be able to handle 9,800A (12v application only) for 3 seconds then end of load. Afterwards, much lower load will follow.
Continuous load- 500-2000A

AGM capability- 600aH

If voltage is lower than, 11.7v, for example, there will not be a 2000A continuous load, it will be lowered.


Current ideas-
Because i would need ~7,300A of surge power from LiFePO4.. I've looked at the following-

240 A123 AHR32113M1UltraBs
120 A123 AHR32157s

48 60aH prismatics (5C continuous, 10C burst)= 720ah, 7,200A surge (0.33C for 5000 cycles) 2.5v shut off
For safety sake, i'd program one of my controllers to kill the entire load if voltage dropped to or below 10.5v with these cells.

My price limitation i'd like to stay under is $4,000.

Those 3 options do that.

I didn't name the brand of the prismatic's above because i do not know it.

I just know it comes inside aluminum case and is considered LY-60aH


My fall back option is to cut the surge load in half from 9800 to 4900 and use less cells (which prismatic any brand would be more open to purchase).

The only thing i'm leary about is this-

If i use 100+ cells, i'm afraid of Uneven Discharge Distribution!

The source of the load only comes from 4 devices, each device with 6 active connections.

So that's 24 connections capable.

With 240 cells with no board, that's 60 packs to connect to.. Not really divisible by 24 and am afraid of uneven discharge rate with so many connections.

Same with 120 cells with no board.. 40 packs on 24 connections..


Would it be safer to stick with prismatic simply due to high amperage in very short usage?
I'm not afraid of the loads not being shut down from low voltage, i'm afraid of very strong discharge rates imbalancing the cells rapidly causing chaos.

I do not know if that's possible on just 1 cycle between manual charging/balancing but i am cautious with taking care of expensive investments.

While the 2nd option of scaling back to 4,900 is a last resort option because more power is always better,

And FYI, i've looked into Ultracapacitors but they will not work for my application because of 2 reasons-
1- too expensive
2- reserve capacity no where near the capability of LiFePO4 once they are used for much lower continuous discharge.

yes, go with that. totally clear what you are doing. it will work fine. not sure how you got 720ah though. most of us use Ah since amps are always capitalized as A.

48 cells-
4S12P
12P x 60aH = 720aH

shizzzon said:

I am easy to deal with in terms of limitations.

Weight- no limitation.
Being used over 10 AGM batteries as typical requirement.. lithium anything is a miracle technology in that sense.

It would be some AGM and the rest an unlimited amount of lithium cells in terms of what is required to run a load.

Sizing- no limitation.
Again, there is plenty of room even for 1000aH winston cells if i were rich, heh.. but seriously, space is not an issue either.

Power/energy-

I have 2 options so i will attack the harder option first to see where i stand.

I need to be able to handle 9,800A (12v application only) for 3 seconds then end of load. Afterwards, much lower load will follow.
Continuous load- 500-2000A

AGM capability- 600aH

If voltage is lower than, 11.7v, for example, there will not be a 2000A continuous load, it will be lowered.


Current ideas-
Because i would need ~7,300A of surge power from LiFePO4.. I've looked at the following-

240 A123 AHR32113M1UltraBs
120 A123 AHR32157s

48 60aH prismatics (5C continuous, 10C burst)= 720ah, 7,200A surge (0.33C for 5000 cycles) 2.5v shut off
For safety sake, i'd program one of my controllers to kill the entire load if voltage dropped to or below 10.5v with these cells.

My price limitation i'd like to stay under is $4,000.

Those 3 options do that.

I didn't name the brand of the prismatic's above because i do not know it.

I just know it comes inside aluminum case and is considered LY-60aH


My fall back option is to cut the surge load in half from 9800 to 4900 and use less cells (which prismatic any brand would be more open to purchase).

The only thing i'm leary about is this-

If i use 100+ cells, i'm afraid of Uneven Discharge Distribution!

The source of the load only comes from 4 devices, each device with 6 active connections.

So that's 24 connections capable.

With 240 cells with no board, that's 60 packs to connect to.. Not really divisible by 24 and am afraid of uneven discharge rate with so many connections.

Same with 120 cells with no board.. 40 packs on 24 connections..

Would it be safer to stick with prismatic simply due to high amperage in very short usage?
I'm not afraid of the loads not being shut down from low voltage, i'm afraid of very strong discharge rates imbalancing the cells rapidly causing chaos.

While the 2nd option of scaling back to 4,900 is a last resort option because more power is always better,

And FYI, i've looked into Ultracapacitors but they will not work for my application because of 2 reasons-
1- too expensive
2- reserve capacity no where near the capability of LiFePO4 once they are used for much lower continuous discharge.

Click to expand...

Would you mind elaborating on what you're using all that current for? Sounds interesting!

Given that you're running a low system voltage, you're only going to have effectively 4 giant cells. As such your issues with balancing are very unlikely to cause issue. Because total capacity is very large to cater for the pulse current the likelihood of substantial drift between these four giant cells is very low. I would not anticipate balancing would be required very frequently. Balancing becomes more critical with higher voltage strings because a single cell failure can lead to overcharging all of the others. This is why there are many stories of LiFePO4 failures on these forums, primarily due to single cell failure due to pack design. While single cell failure will of course have an impact in your case, it's unlikely to be fatal to the other cells quickly.

I'm curious too what your hobby is. A huge tesla coil?

As for lifespan of Lifepo4. Several things have caused problems for those of us here who's hobby is electric bikes. The weight, and cost both tend to conspire together to lead many to choose the smallest possible battery. Then long duration discharges are done at pretty close to the cells max c rate.

I think failure rate can increase from this kind of use. It just puts maximum strain on any cell that is weak. The likelihood of a weak cell may be much more for lithium of all types, than for lead. It's just a lot more complicated a manufacturing process than lead.

Then you have rough handling. The bikes ride on rough roads, solder breaks on BMS wires, pouch cells get dinged, spot welds pop. For bikes at least, a great deal of the failure rate is from mechanical damage of one kind or another.

But for sure, It sounds like you do a large amps, but shorter duration type discharge. In general, Lead can take that kind of punch a lot better than lifepo4. Lifepo4, even the best, is really happiest when doing a steady drain that is half the c rating or less.

Yes I agree.

The Good Group 31 AGM batteries can handle upwards of around 700A for several seconds while still staying around 10.2-10.6v for 15-30 seconds straight.

It's good and all but doing that to AGM will definitely destroy their life compared to lifepo4 assuming i stay away from over 80% DoD.

The hobby is purely MUSIC.


Tesla coil was going on the right track, heh.

Purely for competition and show.
Mainly for show.

I do not know if this interest anyone here but there are starting to become some Car Audio LiFePO4 batteries hitting the market that excel in burst ratings with a price..

One famous company in the AGM business recently released their LiFePO4 battery-

8v cut off
15v max charge (14.5v recommended)

Continuous discharge- 1,080A (12v battery)
Burst rating- 2,400A
540A max for charging

675 watt hours

M8 bolt posts
Full active onboard protection

Price- $3,000

I'm someone who likes to do everything myself.

The lithium market is getting more and more popular in this hobby especially in smaller vehicles where trying to fit multiple batteries isn't possible.

Electric Cali-Mexi, hydraulic car jumper with music kind of Hobby?

If it's high-discharge LiFePO4 then it'll be repackaged (and marked up) A123 cells. There's just nothing else out there commercially. All other LiFePO4 is low-discharge.

There's a lot of high-discharge LiCo about, but four cells in series makes for a 16.8V system (14.8V nominal) so a 12V system would require some modification to work with it.

Punx0r said:

If it's high-discharge LiFePO4 then it'll be repackaged (and marked up) A123 cells. There's just nothing else out there commercially. All other LiFePO4 is low-discharge.

There's a lot of high-discharge LiCo about, but four cells in series makes for a 16.8V system (14.8V nominal) so a 12V system would require some modification to work with it.

Click to expand...

That's exactly what it is. They mention nanotech in their batt description.

No hydraulics at these shows although they used to tie them in a long time ago.

While the majority of events have the music inside the vehicle, sometimes we just have to go crazy-

https://www.youtube.com/watch?v=cWTCI5_pwHU

I'm confused. Your posts are pretty cryptic, but you seem to be into car SPL competition. If so, why are you looking for a 8 or 10 year lifespan from the batteries to power your amps? Also, you say cost is not an restriction, only weight, but then say ultracaps are too expensive.

I would be looking at A123 for the 12V-compatibility of LiFePO4. Possibly salvaged modules, like the Hymotion bus packs, where you already have large numbers of cells welded in parallel. Use a BMS to charge and balance, but run naked for discharge because the power level is so high.

Punx0r said:

I would be looking at A123 for the 12V-compatibility of LiFePO4. Possibly salvaged modules, like the Hymotion bus packs, where you already have large numbers of cells welded in parallel. Use a BMS to charge and balance, but run naked for discharge because the power level is so high.

Click to expand...

+1

If you are looking at LiFePO4, A123 AMP20 cells are your victim. There are a couple of good kits out there to help with assembly, I just ordered and assembled a 4s2p pack (12v nom, 40ah). I'm pretty happy with it.

If you are willing to look at other Lithium chemistries, I'd be looking at getting a Leaf pack and re-assembling it to 4s(whatever)p...

AAHH my ears! thats crazy, movin the earth.youre poor ears :lol: .

Punx0r said:

I'm confused. Your posts are pretty cryptic, but you seem to be into car SPL competition. If so, why are you looking for a 8 or 10 year lifespan from the batteries to power your amps? Also, you say cost is not an restriction, only weight, but then say ultracaps are too expensive.

I would be looking at A123 for the 12V-compatibility of LiFePO4. Possibly salvaged modules, like the Hymotion bus packs, where you already have large numbers of cells welded in parallel. Use a BMS to charge and balance, but run naked for discharge because the power level is so high.

Click to expand...

Long life span = given some abuse here n there will shorten it's life but longer life should still provide longer life than shorter life cells being abused as well. (abused meaning running way over standard C discharge)

Ultracaps are too expensive because they have no reserve power to them. That's not their chemistry. I'd need probably 10,000 of them and that's not realistic. Not for sudden discharge but for longevity use on lower power.

That was my idea on the charge/discharge.

Discharge freely and manually charge them + balancing outside of installation.

Ok, let me look at these options again from everyone elses posts as well and see what it looks like for me.

Thanks for unanimous decisions on A123. I know they have insane discharge ratings.

beast775 said:

AAHH my ears! thats crazy, movin the earth.youre poor ears :lol: .

Click to expand...

As ironic as it sounds from me talking about loud music, i personally like it for maybe 10 seconds.. then i'm done,

It's great feeling that you get nowhere else. Also being able to use it for fun at shows, etc

For actual listening, it's turned way down. I have to be able to see out of the windshield so it's not turned up to ridiculous levels.
I don't want the National Guard called on us.


To change the subject real quick, what is the general Location of members here on ES?

Is the members just global or is the member list here mainly one section of the world?

Most forums i tend have a primary location with some members here n there. Just curious.

Primarly North America, secondly Europe including Russia, Australia incuding NZ.
I wonder if anyone is in Africa.

The power/sound levels are impressive...but I hope it was the recording mic that messed up the sound on that video, because something was seriously distorting ..either mic, speakers , or amps.

I'm still in clear of your requirements, you mention 9800 A for a few seconds, but later you say you need 7,300A burst ??
..and , I guess you are fixed at 12 volts due to other existing components ?

YEs fixed at 12v.

9800A burst total draw from devices.

7300A comes from the requirement from Lithium alone. The other 2500 comes from AGM batteries.

There is a light mix of both technologies for even weight distribution and some lee-way with voltage swing.

They will be separated when vehicle is off so no worries there.

And about distortion, yes, it's the mic.

99.9% of the time, all videos of loud audio/music is distorted.

There are some microphones on the market that attack this problem but i've never owned any.

7k-9k amps, wow!!!!!!
Thats some thick wire!

markz said:

7k-9k amps, wow!!!!!!
Thats some thick wire!

Click to expand...

Ha!

You aren't kidding!

When i was suggested here to wire the 48 prismatics(if i went with those) as 4S12P.. i almost fell out of my chair because that immediately told me i would need some ridiculous wire gauge.

Doing some calculations, i came up with-

Buss bars on them would be good for ~900A-2500A continuous for intervals of 10min and less probably.
A burst of close to 10,000A is fine because it usually last less than 5 seconds.

Now... what concerned me is the series wiring!

There is some audio power wire that is actually extremely flexible 350MCM wire even though it's publicly rated at 4/0. (audio wire is always rated smaller than welding cable for some reason.
IE- 1/0 audio = 2/0 welding. 2/0 audio = 4/0 welding, etc)

Right now, it looks like i would need/want 2-3 350MCM cables to series each 12pack bank up...

I haven't finalized what cells i wish to go with yet as i have some research to do but cabling itself is no joke.

Let me edit and say this because after looking everything over.. it's not as bad..

I'm not used to paralleling cells then seriesing.

I've always been used to seriesing items by themselves to other independent devices.

With 4 12 packs of cells, i can series off of ALL 12 cells rather than just the 1 cell..

The info above is if i had to series off of just one..

Whew! takes a great burden off, hehe.

So, 12 wires to series per pack is a LOT better for my peace of mind.

I'd probably still do standard 4/0 cable.. all 12 runs per pack, just keep them within 4-6" long per cable.

Does anybody know what the ACTUAL discharge C ratings are for A123 Cells and how do i know i'm getting the correct ones?


I just did a search and all providers give different values.

This is no lie, there are all the values i was given from different providers-

AMP20-

5C / 20C
10C / 30C
20C / 40C
30C / 60C

Are they all the same or this there purely fake knock-offs?

they are pure knockoffs. no sin involved.

the 5C can become 20C if it has 4X the effort. 10C can become 30C by trying three times as hard, all of them saved by god's pure grace and no aH wasted. when you have need of Ah then no aH will ever do.

http://www.a123systems.com/
http://www.a123systems.com/prismatic-cell-amp20.htm
click on "Data Sheet" pdf file