Safe LiFePo type lithium vs. Lithium Ion cells

[JoeP]

What would the efficiency loss be if manufacturers substituted safe LiFePo type lithium cells for Lithium Ion cells?

http://www.nytimes.com/2016/09/03/techn ... ticle&_r=0

http://www.nytimes.com/2016/09/03/busin ... pe=article

http://www.nytimes.com/2016/09/09/busin ... pe=article

http://www.nytimes.com/2016/09/13/busin ... pe=article

[Burgerman]

They would need about 2 times as much battery volume. Makes the phone bigger. Or have half the run time. They would get a much longer service life and huge safety improvement.

Same with the tesla cars. They offer huge 270 mile range and 150mph. (not both!) because they use lithium ion.
If they used lithium ion phosphate then they would need a battery double the physical size. Which will not fit, and costs much more. Again it will last longer.

[Burgerman]

Chemistry, performance, cost and safety characteristics vary across LIB types. Handheld electronics mostly use LIBs based on lithium cobalt oxide (LiCoO
2), which offers high energy density, but presents safety risks, especially when damaged.

Almost all consumer stuff, and tesla cars. Boings dreamliner, cameras, phones, portable powerchairs, everything you hear about in the news that burns or explodes. Boom... Not listed, LiPo as used in my planes and quadcopter. Much worse still! And the most powerful by size and weight of the lot by far.

Lithium iron phosphate (LiFePO4), or lithium ion manganese oxide battery (LiMn2O4, Li2MnO3, or LMO) and lithium nickel manganese cobalt oxide (LiNiMnCoO2 or NMC) offer lower energy density, but longer lives and inherent safety. Such batteries are widely used for electric tools, medical equipment and other roles. NMC in particular is a leading contender for automotive applications.

What we use... LiFePO4 by far the most common in this list. Light, long life, doesent burn, lower C rates, physically bulky.

Lithium nickel cobalt aluminum oxide (LiNiCoAlO2 or NCA) and lithium titanate (Li4Ti5O12 or LTO) are specialty designs aimed at particular niche roles. The new lithium sulfur batteries promise the highest performance-to-weight ratio.

Unknown/unobtainable variants... Experimental mostly.

https://en.wikipedia.org/wiki/Lithium-ion_battery tech

[JoeP]

I read somewhere that Boeing solution to its combustible lithium battery design was to put it in a fireproof box! That seems like engineering malpractice to me.

[Burgerman]

It is. The problem was as usual, a BMS induced overcharge due to failure or cell imbalance resulting in a fire in several cases. I could have predicted that without even seeing it. But everybody KNOWS that you gotta fit a BMS!!!

[JoeP]

Why is it so difficult to design a battery management system that actually works and doesn't melt-down?
Boeing & Lockheed Martin are chockablock with engineering geniuses. Heck, they joint venture to launch rockets under Launch Alliance. With all that engineering brainpower you would think that a BMS that actually works would have been designed by now.

[LROBBINS]

Design and production of the Dreamliner battery was contracted out to Yuasa. They too should have had the engineering savvy to avoid their catching fire.

[Burgerman]

If you look carefully in the images of the burned battery you can see a generic BMS that looks just like the typical ones sold by the likes of evassemble...

They think that the charger input power on/off MOSFET failed on that BMS, I read somewhere.
Meaning that the charger kept going, until it was finished charging. And the BMS was unable to disconnect the charger Amps. Not a problem, normally on say a lead battery, because the charger voltage was 29.4V and CORRECT for a 7S pack lithium ion pack. (I am presuming it was a 24V battery). With a 4.20V charge per cell, that's the CORRECT voltage.

But the issue is that on charge, towards the end, there's always 1 or 2 cells that shoot up to the 4.2V (full) limit long before the rest do. What normally happens is that the BMS applies its feeble 100mA 'brake' or a pulldown current, which does almost nothing because the charger is maybe pushing in 10A or 20A... So the BMS now sees the volts exceed 4.25V or similar, and so (as per BMS normal behaviour), DISCONNECTS the charger for a bit, until the HIGH cells weedy balance current can lower the volts to around 4.15V...

Then once accomplished, it allows full charge power in again by reconnecting the dumb charger, and so the cell goes high, (again and again for hours) then rinse and repeat on/off until the low cells catch up. Once ALL are at 29.4V (4.2v per cell) no more balancing or charging happens. But the failed mosfet (fused on) couldn't turn off the chargers Amps supply... So those high cells went well over voltage. Maybe a lot over. Do you know what happens to a lithium ion cell when its voltage goes a tenth of a volt or more too high? Yep, fire!

[c500user]

Why not make a BMS that behaves like a battery charger? If it is safe to charge LiFePO4 using a PL8, surely it must be safe to charge using a PL8-like BMS.
The PL8 design could be cut-down to perform only the required functions. Although it might not be cheap, it would be a safe and should be smaller and cheaper to build than a normal PL8.

Note that Victron seem to have understood with their Peak Power Pack product. A 12V battery that is based on LiFePO4 cells with an integrated charger. Not cheap (12.8V 40Ah € 900 + VAT) and perhaps not really suitable for wheelchair use, but maybe the idea will catch on.
https://www.victronenergy.com/batteries/peak-power-pack
https://www.victronenergy.com/upload/do ... ack-EN.pdf

[Burgerman]

I imagine those are dumb charger + bms + battery.

[c500user]

I think you assume wrong. Input can be 11-25V (to charge from car, solar or whatever). Sounds like more than a BMS to me.

[Burgerman]

Its quite possible to design a full system, that has correctly designed components, motors/battery/controller/charger to work correctly as one well thought out system. As most EV cars/bikes already are.

What lithium isn't is a battery (with its built in BMS), as a separate thing on a shelf, that you just put into various user chosen devices. As lead bricks are today. Where the thing it is to be used on, and the charger, all the settings, and the device settings and current requirements is unknown. Every part needs to be chosen and configured to work together as a part of a system.

The reason we don't have lithium in powerchairs is that the manufacturers don't understand this. They are like mark on the other site. Waiting for a lithium version of the universal square 12v brick they can just drop in. They are waiting for a lead brick replacement. After 10 years of this waiting they are STILL waiting. When the manufacturers make a dumbed down brick to do just that for dummies it will still be a compromise. Because to take full advantage the chair needs to change, the voltage needs to change etc.

[c500user]

Well, I do not fully agree. What you advocate is bespoke. I think building blocks can work, if chosen correctly. Awell thought out design can be used for any purpose that fits within the design specs. I agree that if I take a component with spec A and use it for usage B it will fail or cause other problems. But that goes for batteries (lead or LiFePO4), motors, etc.

If you look at the Victron website and datasheet, it is designed for use in Caravan movers. These require short bursts of high-amp power (when they are moving the caravan to the desired position). Output current is specified as 150A continuous, 200A 10 sec, even for the smallest 8Ah unit.
Most of the time they are left to sit there till they are needed again. Given the average age (and technical know-how) of caravan users they need to be fool proof.

From what I gather from the datasheet it seems a well thought out product that can be charged in many ways (if only our wheelchairs would be charged so easily without modification).

[Burgerman]

But for a bunch of reasons there isn't room for, their solution has exactly the same issues as the lead brick replacements we can already buy for wheelchairs. That do not offer reliability, better lifespan, or greater range. And so it is a far from optimal solution, and not one what will work for a powerchair. And for the reasons I already gave here: viewtopic.php?f=2&t=1813


EG. There's hundreds of people already offering lead brick replacements. They mIss out on all of the advantages of actually going lithium, and most wont actually work due to the BMS protection system being incorrect for your application or because of low C rate cells..

Here's a typical 100Ah lithium ion phosphate battery. Ready to go! This one is physically too big for 100Ah - they all are...
If you find one that will fit (grp 24 sized) it will be 50Ah high rate or 70Ah low rate at best. And it wont work for many reasons, even if it did... Same exact issues with the Victron. And will not give you better range, or longer life than the same physically sized lead. So you would be swapping to lithium for no or little gain, at great expense!

[Burgerman]

Well, I do not fully agree. What you advocate is bespoke. I think building blocks can work, if chosen correctly. Awell thought out design can be used for any purpose that fits within the design specs. I agree that if I take a component with spec A and use it for usage B it will fail or cause other problems. But that goes for batteries (lead or LiFePO4), motors, etc.

No it doesent! And that's the problem. A caravan doesn't take intermittent 200A to 240A loads like a R-net 120 can, and doesn't care about Ah per volume, or C rates etc. And even ignoring the BMS issues, the RATE you discharge or charge a Lithium cell, has a truly massive affect of its cycle life. As does the voltage you charge to. Or the voltage you discharge to. So fitting a too small badly packaged brick means you may get 2 or 3 years rather than the 10 plus that is available done correctly. If the BMS doesn't damage it sooner! So for 1400 dollars you get basically the same as the MKs... No sense at all!

EG Straight swap for your lead to lithium. And its got a high amp BMS too. (but still wont work on a steep ramp for eg it will cut power to protect the overworked battery). It will give the same range as your MK. And probably might last as long/longer if nothing is mismatched or fails. But no real advantage gained. http://www.lithiumion-batteries.com/pro ... n-battery/

All you gained was expence and unreliability of BMS going into an environment it doesn't understand or care about. Its job is to stop your chair taking too much from the small battery. It entirely misses the point of going lithium and the advantages it CAN offer!
EG, my BM2 battery is also grp24 sized, but uses 10C cells for less damage from pulses of high rate discharge, and is 120Ah further reducing these pulses by 2.5x per cell. Nandols is 190Ah if he ever collects it. This means we don't work them hard, and they are high rate cells anyway. So the 240A we intermittently take don't worry them, hence very long life. And proper charge control = long life. And also 2.5x the range of the battery above... And so less charge cycles required every week, or lower discharge levels, meaning longer life... Again. We win 4x by doing it correctly.

A bms is basically to allow a battery that's too small, to be fitted somewhere that it isn't suited to. It isn't required at all in a properly implemented system, so wont keep cutting off power as you try and drive the chair. So the real answer is to match the cell type, capacity, to its demand, let the controller take charge of current limiting and do so proportionally as it already does, and charge with a hobby style charger matched to use and pack demands.

[Burgerman]

You may have noticed that almost every lithium battery system IS bespoke! EV cars and bikes. Drills. Portable powerchairs, ev bicycles. even your laptop.

Well, I do not fully agree. What you advocate is bespoke. I think building blocks can work, if chosen correctly.

If chosen and built completely correctly, it becomes bespoke... The headways ARE building blocks. And that's why I use them. In my case 13S at around 45 to 43V working for best efficiency.

What you advocate is bespoke.

Because they are a part of a bigger more complex proper management system that it needs to fit! There may be computer controlled cell balancing going on in an EV for eg, with partial top up charges on the move, or specific charge algos to match a cell type or temperature/age with sensors on different parts of a pack on an EV car for eg. The controller may allow different peak amp draw by current limiting the controller pulsewidth in use, determined by exact pack parameters plus age, or cell temperature, or measured internal resistance or state of charge. Or all of the above, to maximize efficiency or cell life. Likewise, it may charge differently depending on many different but fine tuned parameters of a specific cell, or pack. These are all critical for best performance, efficiency, service life etc. Some EV batteries for eg include heaters to get best performance and no damage in cold conditions based on sensors built into the packs. And different companies are all using different variations on exact chemistries, etc.

The idea of random lead brick replacements sat on a shelf, that are not sensitive to or optimised for, without any idea of the system they will be going into, is really a non starter other than what we already have designed for generic use with a anti dummy BMS built in to try and stop instant damage.

Here's ONE reason its the wrong way to do it. All EVs or even mowers and bikes, use much higher voltages. Because it allows smaller Amp controllers, much higher efficiency, and smaller cheaper lighter batteries to do the same job with the same performance or range as before. This isn't possible with lead for a few technical reasons. So a correctly designed chair would be maybe 12Ah, 96V, brushless, and have huge range, and much lighter cheaper motors, wiring, etc than before. And no longer need to be built like a railway engine to carry two 60lb lead bricks about, further increasing efficiency.

The idea of dropping two universal lead brick replacements in will always be inferior, cost more, work far less well, and miss out on all the advantages lithium offeres done correctly.

What I am doing is half way in between. Because I cant make or find suitable brushless motors or controllers to do what I really know is required. Fitting an on board, stripped down, PL8 as a dumb charger is what a very basic EV car already does. On a powerchair we lack space. That PL8 space is better used for battery volume. And leave it off board. Where its PC control and configuration and cell monitoring is also useful. Ask Expresso. He wouldn't want a dumb charger now! Same with the BMS and the grp 24 'case'. It just ends up with lower capacity and problems.

[Scooterman]

NMC cells? A viable slightly cheaper alternative to lifepo4?

Discuss

https://m.alibaba.com/product/62001367650/rechargeable-lithium-ion-prismatic-cells-3.html?s=p&spm=custom.wap_new_search.1998817009.5.10db284fQb9dQg

[Burgerman]

https://batteryuniversity.com/learn/art ... ithium_ion

They are a minor variation on a lithium ion laptop battery chemiatry so can be thought of as the same basically..
300 to 700 cycles.
2000 cycles if only 70% of capacity is utilised, but then the main advantage is lost.
Slightly lower C rate and greater capacity than the typical equivelent sized lithum ion battery. But not enough to notice.
They are 3.7V so 7S needed for 24v replacement.
Very energy dense like the tesla cars use, your phone uses etc.
But they burn or explode violently if you get anything wrong or the manufacturer does...
So not for disability use as we need non exploding batteries.