Lithium mini batteries

[rabbi]

Real life obviously beats lab & ABYC test when it comes to "real life conditions"

I think we agree to disagree then. I will continue to consider my LFPs a powerful and somewhat dangerous beast and treat them nicely so it doesn't bite me.

[s/v Jedi]

Actual testing with evidence

https://www.youtube.com/watch?v=Qzt9RZ0FQyM

[CaptainRivet]

Originally Posted by rabbi

The thumbnails I forgot to attach:

If you wanna do that to your batteries use LTO, you can even drill into them while using them :-)
https://youtu.be/eAUYbSDEy6I?si=DehQPRhaqsGYi-_k

[CaptainRivet]

Originally Posted by carstenb

Not sure how your boat is set up, but on ours we can't run the anchor winch unless the engine is running so there is no real strain on the batteries. We do have a bow thruster, but use it so rarely that I simply can't see that this should be a problem.

Again - LiTime warranties the batteries for 4000 cycles so that is over 10 years. More than enough for us

Yes my windlass works without engine on my 1088AH bank without any issue, again a saftey aspect i can still lift the anchor if the engine doesn't start due to any reason. I don't need an alternator to support a weak lead to be able to handle the windlass for a longer periode of time if you eg have to hoist 60m of my 80 m chain and my solar will deliver more during day then my alternator will.

Did you check if your bowtruster is even allowed to be connected to LFP? Some rely on the high resistance and voltage sack of a lead to protect the motor which is stated in the installation manual and warranty doc. And if you connect such a bowtruster to LFP the motor will get damaged by overload. 2nd the bowtruster create huge surge loads which are an continuous danger to the mosfets in the mosfet BMS found in li-times and many other drop ins. To be able to operate the bowtruster savely with mosfet BMS on LFP the continuous current rating of the BMS has to be equal or higher then the surge current you measure or stated in the installation manual of the bowtruster (is rarely there).
And this is not equal to the CCA rating of lead they spec to be used.

[carstenb]

Originally Posted by CaptainRivet

Yes my windlass works without engine on my 1088AH bank without any issue, again a saftey aspect i can still lift the anchor if the engine doesn't start due to any reason. I don't need an alternator to support a weak lead to be able to handle the windlass for a longer periode of time if you eg have to hoist 60m of my 80 m chain and my solar will deliver more during day then my alternator will.

Did you check if your bowtruster is even allowed to be connected to LFP? Some rely on the high resistance and voltage sack of a lead to protect the motor which is stated in the installation manual and warranty doc. And if you connect such a bowtruster to LFP the motor will get damaged by overload. 2nd the bowtruster create huge surge loads which are an continuous danger to the mosfets in the mosfet BMS found in li-times and many other drop ins. To be able to operate the bowtruster savely with mosfet BMS on LFP the continuous current rating of the BMS has to be equal or higher then the surge current you measure or stated in the installation manual.

I'm fully aware how my batteries are set up. We virtually never use the bow thruster - when we do the engine is running we have a 160AH generator on the engine

I simply can't get up enough concern to worry

[rabbi]

Originally Posted by s/v Jedi

Actual testing with evidence

This lab test is proudly presented by GWL, distributor of LFP cells.
I call this marketing.


But I don't see the point to watch videos that show LFP is 100% safe in lab tests while we have (very rare!) real life examples that show fire & explosion is a possibility.

The proof is in the pudding. For me real life beats lab tests.

[CaptainRivet]

Originally Posted by carstenb

I'm fully aware how my batteries are set up. We virtually never use the bow thruster - when we do the engine is running we have a 160AH generator on the engine

I simply can't get up enough concern to worry

Did you read my post?

It has nothing to do with surge of a bowtruster if a 160A alternator is running in parallel or Not, an alternator cannot provide this short-term surge current, so it hits the mosfets of the BMS with the total surge current and because the mosfets are not calibrated there is always one weakest mosfet who gets the full surge current for the first milliseconds...its not one mosfet its typically min 4 that share the load, well not in the first milliseconds.

To avoid all this problems i use a non current carrying eletrodacus BMS on my hybrid starter&house bank means the surge current my bank can deliver is only limited by the peak current delivery of my cells, the fuse and cable it is connect to. and the bank/cells can do 3000A peak, 6 times higher then the highest surge current i ever measured on my vessel

[CaptainRivet]

Originally Posted by rabbi

This lab test is proudly presented by GWL, distributor of LFP cells.
I call this marketing.


But I don't see the point to watch videos that show LFP is 100% safe in lab tests while we have (very rare!) real life examples that show fire & explosion is a possibility.

The proof is in the pudding. For me real life beats lab tests.

Yes real life stupidity knows no borders like putting a cat into a microwave or the typical US bs on a coffee cup warning "beverage is hot...." yes i buy a hot coffee, its freshly brewed upfront my eyes so i can expect it to be hot when i receive it and its my own stupidity if i burn my tongue drinking it right away without caution...
Well not so due to US court where everything can happen...

But by following given installation recommendations and common sense what GWL tested is real life safety.

And that you only have rarely these real life examples is a proof that a lot of things must have gone completely wrong before! that this can happen.

The meanwhile well known nearly daily happening Tesla ION battery fires but non with Lifepo4 cells burned is one off the proofs. And in EVs the cells are operated close to their speced limits and speed charged at their max charge rate. While in boats its exactly the opposite you have large capacity banks 400-1500AH typically that see 10-20A constant and tops of 200-300A when electric cooking or using the windlass for longer to hoist 60m of chain. And charging due to limited space of solar space and small engines that can support a max 200A alternator you are at 0.3C charge rate at best...a lot are far below.

[rabbi]

I fail to see any stupidity that caused the two examples above.

The guy looking for new house is a German electrical engineer. Official result of investigation is that the cells developed a short.

[CaptainRivet]

Originally Posted by rabbi

Electrical shorts are not unheard of in batteries so while not exactly self-combustion the end result is the same: A burning battery.

I have no reason to believe that the cruiser with the burned cell faked the story. But with no video evidence others may choose to consider this fake or whatever.



Same for the other story I mentioned: Homeowners were alerted by an ordinary smoke detector in the basement, found one of their drop-in batteries was smoking heavily. Minutes later an explosion caused the building to partly collapse.

The owner was prettty open about the findings of insurance investigations:
One of three 48v drop-ins with pouch cells started to gas during normal operation, according to the VRM logs at around 60% SoC with charge current below 0.3c and no obvious voltage spike.
The gas exploded due to some electrical spark. They could not identify the reason why the cell started to gas.
Plenty of news coverage, forum entries, etc so I have no reason to believe this is fake.
https://youtu.be/SrLdBvcheqg?t=18



All I'm saying: While LFP has a good safety record it is not 100% safe. It is powerful and sometimes sh*t happens.

The nearly 100% safe chemistry is LTO.

2nd yes i hear more and more about pooch cells creating gas explosion, won't happen with prismatic lifepo4 cells.
The dropin 48V pooch cell was gasing at 60% SoC means it was heavily out of balance and serveral pooch cells got overcharged, eletrolyt gased out and created a internal cell short when getting dry (like an AGM) and gas created an explosive atmosphere. It was a chinese surprise box without proper external communication so you couldn't see whats going on inside the battery and if cells are peaking. The issue here is 48V and 16S adding tolerances and being able to quickly overcharge a cell.


The next issue is 48V means 16 cells in series means 16 times added tolerances.
We already tested that 10 years ago and the austrian NPO that produces the LTO titan1 did the same:
12V or 1P4S are so stable you don't need a BMS IF you get high quality cells means Yinglong 40AH LTO or winston cells or highest grade CALB and even the new V3 version EVE. Proper top balance and they stay in balance and simply work....installed min 100 in cars which are running up to 17 years now without a BMS, LTO just an active balancer. Winston is even selling these drop in 1p4S 12V in 40/50/90 and 200AH without! BMS since more then 10 years.

24V or 1p8S are partly stable and need a BMS with a small balancer (active or passive), even the highest grade cells get out of balance over time as also the distance from first to last terminal creates a different resistance then 1st to 2nd. You can in emergency run the 24V battery without BMS for a limited amount of time, i tested 8 weeks LTO and Calb and austrian NPO 12 weeks with LTO, Winston and EVE till it was serious out of balance.

48V or 16S are unstable and need a BMS with a sufficent balancer, passive balancer only work with the highest quality of cells. the lower the quality of cells the more powerful the active balancer must be to keep that under control. A 48V battery can only be operated without a BMS and balancer with a very limited amount of time. The austrian NPO did ended up developing an own BMS with very powerfull active balancer for their Titan1 powerwall with LTO due to these test.

Another finding was low quality b-grade and C-grade junk cells that worked in 4S and 8S cause massive issues in 48V/16S configs. To make it then even worse they do setups in 3p16S or 4p16S with cells which already not sufficent for 1p16S.
And thats what i see in the forums of offgrid 16S or even 32 in a row compressed together that even minimal expension and contraction due to temp causes mechanical issues on the terminals leading to internal short with pooch cells over time, pprismatic it depends if they are coiled around the anode and catode or layered up in z form have a much lower risk getting a short to pooch.
This is ok up to 8 cells in a row, more you need flexible busbars.
It shows again it needs a chain of events and the more cells in series the more can and will go wrong.
Whats ok and works fine in 4S and 8S config, won't anymore in 16S.

[rabbi]

Originally Posted by CaptainRivet

The nearly 100% safe chemistry is LTO.

2nd yes i hear more and more about pooch cells creating gas explosion, won't happen with prismatic lifepo4 cells.
The dropin 48V pooch cell was gasing at 60% SoC means it was heavily out of balance and serveral pooch cells got overcharged, eletrolyt gased out and created a internal cell short when getting dry (like an AGM) and gas created an explosive atmosphere. It was a chinese surprise box without proper external communication so you couldn't see whats going on inside the battery and if cells are peaking. The issue here is 48V and 16S adding tolerances and being able to quickly overcharge a cell.


The next issue is 48V means 16 cells in series means 16 times added tolerances.
We already tested that 10 years ago and the austrian NPO that produces the LTO titan1 did the same:
12V or 1P4S are so stable you don't need a BMS IF you get high quality cells means Yinglong 40AH LTO or winston cells or highest grade CALB and even the new V3 version EVE. Proper top balance and they stay in balance and simply work....installed min 100 in cars which are running up to 17 years now without a BMS, LTO just an active balancer. Winston is even selling these drop in 1p4S 12V in 40/50/90 and 200AH without! BMS since more then 10 years.

24V or 1p8S are partly stable and need a BMS with a small balancer (active or passive), even the highest grade cells get out of balance over time as also the distance from first to last terminal creates a different resistance then 1st to 2nd. You can in emergency run the 24V battery without BMS for a limited amount of time, i tested 8 weeks LTO and Calb and austrian NPO 12 weeks with LTO, Winston and EVE till it was serious out of balance.

48V or 16S are unstable and need a BMS with a sufficent balancer, passive balancer only work with the highest quality of cells. the lower the quality of cells the more powerful the active balancer must be to keep that under control. A 48V battery can only be operated without a BMS and balancer with a very limited amount of time. The austrian NPO did ended up developing an own BMS with very powerfull active balancer for their Titan1 powerwall with LTO due to these test.

Another finding was low quality b-grade and C-grade junk cells that worked in 4S and 8S cause massive issues in 48V/16S configs. To make it then even worse they do setups in 3p16S or 4p16S with cells which already not sufficent for 1p16S.
And thats what i see in the forums of offgrid 16S or even 32 in a row compressed together that even minimal expension and contraction due to temp causes mechanical issues on the terminals leading to internal short with pooch cells over time, pprismatic it depends if they are coiled around the anode and catode or layered up in z form have a much lower risk getting a short to pooch.
This is ok up to 8 cells in a row, more you need flexible busbars.
It shows again it needs a chain of events and the more cells in series the more can and will go wrong.
Whats ok and works fine in 4S and 8S config, won't anymore in 16S.

I fail to see a chain of events that lead to failure of the 16s, as there was no chain of events. It was a drop in with it's own BMS that failed during normal operation.

The only thing known is that all of a sudden one battery pack started to gas and minutes later the gas exploded.

Is LFP relatively safe technology? I think so.
Is it 100% safe? Nope.

[CaptainRivet]

Originally Posted by rabbi

I fail to see a chain of events that lead to failure of the 16s, as there was no chain of events. It was a drop in with it's own BMS that failed during normal operation.

The only thing known is that all of a sudden one battery pack started to gas and minutes later the gas exploded.

Is LFP relatively safe technology? I think so.
Is it 100% safe? Nope.

I explained why and how it failed, a combination of bad BMS,Pooch cells, 16S config and drop in surprise box.

wouldn't have happened with
A) prismatic cells where the material is wrapped around Catode and Anode like Lishen does. Has the disadvantage that you only get 272AH into the same case EVE puts 304AH folded in Z style. But has the advantage safest Lifepo4 cell design.
B)in a 4S or 8S config
C)independent proper last resort cutoff

[SV Confianza]

Originally Posted by CaptainRivet

Thats a great approach.
How many watts does it have 700 or 1000W?
Normal opperation of rhe windlass 6 in parallel yes but was the anchor stuck during one of these pulls?
My 1000W showes also draw of around 100A but i saw 300A on the BMS stats and surge was for sure double when my anchor was stuck underneath a rock.
yes 6 in parallel will handle that, its a question of as i described of the mosfets of the BMS.

Actually its spread over 1000ah. I have 6 100Ah minis, and one single LiTime 400Ah, all in parallel. All different length runs between batteries. No bus bar. Basically doing everything wrong. A few class T fuses in between for protection.

Trying to do a long term test to see how much it really affects these cheap batteries and how much imbalance I have while charging.

I have had to lower my charging voltage to keep the one on the end from overcharging. But let’s see how this mess holds up after a year. I have my older battleborns sitting at home so I can test these instead.

[CaptainRivet]

Originally Posted by SV Confianza

Actually its spread over 1000ah. I have 6 100Ah minis, and one single LiTime 400Ah, all in parallel. All different length runs between batteries. No bus bar. Basically doing everything wrong. A few class T fuses in between for protection.

Trying to do a long term test to see how much it really affects these cheap batteries and how much imbalance I have while charging.

I have had to lower my charging voltage to keep the one on the end from overcharging. But let’s see how this mess holds up after a year. I have my older battleborns sitting at home so I can test these instead.

:-) i like that...
Well why not at one busbar and connect them properly because like this you are not really testing them just that the cable mess might work or not...
They are ok quality and worth the money, the ones with prismatic cells..simple chinese surprise boxes and if installed well they work reasonably well. Just avoid the ones with parallel or pooch cells inside.
...not the pooch cells ones as they have several safety issues in general. One of them if overcharged the eletrolyt turns into emit a gas that can explode and if all electrolyt is out it creates an internsl cell short that actually creates the spark that triggers the explosion.
Pooch cells are the AGMs of lithium with the advantage for AGM that the AGM gas emited is only toxic but not explosive...perfect combo get a AGM starter and pooch cell house and charge with defective regulator on your alternator...for the ones with a death sentence wish...

Both is what i don't want on my vessel....
Would put battelborn and the 400AH litime, install correctly and you are good for a long time.

[Letterkenny]

I’m happy with my 4x litime minis. I’m space constrained so these worked perfectly for me. If I could have fit Epoch I would have but I can’t justify battleborn prices.