Commercial lithium batteries?

[JoeC]

Hi all! Sorry I’ve been away for a while, I hope you’re all doing well.

I’m back because I’ve recently started learning a lot more about lithium batteries, and although I haven’t touched a wheelchair in a while I can’t believe how much better lithium is than the gel and AGM batteries that everyone is stuck with. If you look at an automotive standard lithium battery, you could get the same energy as two lead batteries stored in about 1/5the the weight, for the price of one battery, and it would last for easily 1000 full cycles and charge in about an hour without any difficulty. Why don’t we have this?

So, I’m asking questions and seeing what it would take. Trying to make some networking inroads on LinkedIn, and maybe I’ll get to talk to the right person. If you’ve exhausted this line of inquiry and don’t think it’s worth it, I understand... but I’d like to see if I can do anything.

[JoeC]

As far as I can tell, one of the biggest barriers is shipping of the batteries and completed chairs, and transportation when you want to fly with your chair. I’m reading this looking for more insight: https://www.ecfr.gov/cgi-bin/retrieveEC ... .12.1.25.5

[greybeard]

Some light bedtime reading for you. viewtopic.php?f=2&t=1813

[JoeC]

I guess that's what I get ¯\_(ツ)_/¯

[Burgerman]

Quick update... Summary.

There are lots of new cells appearing.

These for e.g. will fit in the place most group 24 lead batteries came out. You need 8.

Because of voltage drop, peukert, etc we get about 40 to 45Ah at best from say MK gels. Before the chair stops. 15 real world miles, 9 months, (25 to 27 claimed ISO miles). With these below you will get 7 to 10 years, 60+ miles real world...

The batteries here https://www.lifepo4-batteries.com/sale- ... f-car.html

will fit, offer 7x the service life, 5 times the range (176 usable div by 40Ah peukert) and saves 50lb. And can be charged at 40A with a lightweight hobby charger. For 2x the price of lead. They will outlive 2 chairs.

And older tech cells, 45V and 3.2kwh. http://www.wheelchairdriver.com/BM-MK3- ... rchair.htm

[JoeC]

Thanks for the summary! Looks like you've moved on from the Hyperion charger to a Powerlab charger too.

The thing I'm looking at is.. what can a commercial operation do? I see one or two chairs out there offering lithium batteries at this point, and I wonder if that's as good as it's going to get. Most car manufacturers are now pouring a huge pile of money and engineers into lithium batteries, and things are going to get better and better. $100/kWh is already done, the next milestone is even cheaper.

[ex-Gooserider]

The problem is that to do Lithium the way that BM does, needs a fancy expensive charger, and a pretty significant amount of technical knowledge / skill on the part of the user... It is seriously NOT 'idiot proof' and a user can quite easily destroy the pack with some simple mistakes....

Users are used to the basic brick chargers that you just plug into the XLR plug and wait for the blinking lights - and the PL8 is MUCH more involved.

The only way to get around this is to use some sort of Battery Management System (BMS) which if not carefully and properly engineered for the specific application ends up being as we refer to it a 'Battery Massacre / Murder System' for reasons BM has explained endlessly in other threads...

The car makers can afford to (and do) some serious specialized engineering to make dedicated BMS systems that allow them to do relatively simple 'plug-in' charging.... But as individual users it isn't practical for us to do this as it is seriously involved - but the 'one size fits all' BMS systems do NOT work at all well for chairs.

Since chairs have different controllers, with different loading, motors, etc. I presume that each would need a different set of programming, as well as needing different parameters for every pack configuration... Given modern microcontrollers, it might be possible to design a chair specific BMS, but it would be much harder to come up with a way to program it....

ex-Gooserider

[JoeC]

In order to get to a solution that's cheap, reliable, plug-and-forget, we need to be thinking on the scale of hundreds of thousands of chairs. This means one of the companies making controllers will need to make a new version that's specifically intended for lithium. It would manage the balancing and health of the batteries. It seems like it would have to live in the chair, either in the controller or the pack, and you'd keep some kind of two or three prong low-voltage DC connector that the users would handle. Certainly not an XLR plug, but... maybe not that different.

I would hope that while they're at it they could move to brushless controllers, but I'm not holding my breath. The main reason that I see hope here is that you could have a battery that costs less than lead acid and could potentially last as long as the medical system deems that a person should have the same chair. Cutting out the battery changing appointments (that US vendors tend to perform at a loss) would make some bean counters happy.

[Burgerman]

I will reply tomorrow. But to get any real benefit from lifepo4 packs you must go big. Not cheap. Ideally replace GRP 24 with the cells I linked to.

[JoeC]

I know it's late there, I look forward to the reply!

In the meanwhile, I'll let you know that what I'm thinking about, something for your morning coffee. What I have in mind is about around 96 pieces of something like this: https://www.orbtronic.com/20700-panasonic-sanyo-ncr20700b-battery-rechargeable-li-ion. Never mind the unit price on that web site, they'll be around $2 each when purchased in the millions. That's 13.2 pounds of cells, roughly $200 cost, occupies a fairly small volume, and has over 40AH of usable capacity, and is able to be charged fairly quickly. Not appropriate or economical for a one-off, but if you have robotic assembly with laser tab welding, punched busbars, and custom injection molded and extruded parts, I think it could be a very attractive step up from gel cells.

[flagman1776]

The cells you linked to are Lithium Ion, not LiFePO4 which are much safer & have many times the cycle life.

[Burgerman]

Theres a reason we all use LiFePO4 cells though. Actually theres 3 very good reasons.

1. They do not burn, or explode. Unlike lithium ion. Lithium ion is cheap, more energy dense, but I for one wouldn't sit on a stack of them. Did you ever see how explosive and how fiery they get when a laptop, or phone or tesla car goes up? I have actually had a lithium ion (cobalt type) cell catch fire on my workbench, and for reasons unknown. Otherwise they are a good cheap source of storage power.

2. Those lithium ion cells if used to full capacity, have around 500 cycles at best in them discharged gently and charged accurately. (Compared to LiFePO4's 1500 to 2000 if used heavily, or 5000+ if you fit a large enough pack so that you dont discharge hard -- high C rates murder lithium, or and avoid deep DOD most of the time.) And the harder you discharge them (by using a small pack) the less you get. Lithium likes to be mollycoddled... If you want a sensible cycle life from that chemistry you must do as Tesla do. They charge to only 85% and discharge to only 30% in normal daily use. In other words, to get a good cycle life, they only use 55 to 60% of the full pack capacity. Their 60kWh pack is actually a 90kWh one. At which point the added energy density or cost over LiFePO4 is no longer significant. Fitting a 40Ah pack in a powerchair sucking hundreds of amps intermittently, and using most of the capacity will see about the same lifespan as my lithium laptop battery. A year at best.

3. Voltage. A fully charged LiFePO4 cell is 3.600V so 8 is 28.8, same as a lead batter setup Which drops fast to 26V. Same as your gel battery. And they maintain around 13V / 26V giving no compatibility problems with mobility equipment all day long. When they begin to drop a light or two, your day is over. Typically 40 or 50 miles. With Lithium ion, you need 7S and the charged voltage of 7 in series is 29.4. And it doesn't drop to a nice level for much of the day. And that gives over voltage errors on some controllers. That could be fixed.

[JoeC]

It makes perfect sense why everyone here is using LiFePO4. For DIY purposes, almost every one of your points is dead on. I'm not so sure about capacity... the Tesla cells (and some other car cells) are tested to maintain 80% of their starting capacity after 1500 cycles. The linked 176AH batteries claim to maintain 60% after 2000 cycles at 80% DoD. Unless you put them through the cycles, I'm not sure you can know that the LiFePO4 cells really do what they claim.

Regardless, I think that the thing stopping widespread adoption of lithium in powerchairs is a business problem, not a technical problem. The methods and materials you use here could be refined for mass production and high reliability for hands-off untrained users, but I'm not sure any existing company would take on the expense to put it through development. On the other hand, the batteries used by car companies are going to be cheap as dirt in a few years. They're fighting hard to get the price down, and they have a strong incentive to invest in safety of pack design. I'm looking at the technical capability of these things and I'm asking, is there a way for a wheelchair company to ride that wave?

[Burgerman]

The fact that they are basically a firework, means that you see rather exiting fires from teslas now and again. In spite of the cells being heated and cooled by water and monitored very closely. And built into huge grids with each cell fused individually in a very clever design to try and prevent any chain reactions getting going. But in a car, that pack sits under the car which is a metal box to allow you to stop and get out. A wheelchair user such as myself doesn't have that option, and we are usually indoors too.

Tesla use a very careful management system, and around 66 percent of the rated capacity to get their endurance. And again they are using a very large capacity pack. The reason they do so is not only to get up to 300 miles per charge, but because theres only one way to get high cycle life from lithium packs.

They will give very high numbers of cycles if not worked hard. And the opposite is also true. So lets say your 120A Rnet chair can take 120A periodically. And 200A frequently when zero turning. Or when ramp/curb climbing. And lets say you only have a 40Ah pack as you propose. That means that your cells will be very heavily loaded (Amps) regularly. The relationship between how much a lithium cell is loaded and its cycle life is almost linear. I use those lithium ion 16550 cells in some hobby stuff. After 100 cycles, charged extremely accurately they are toast. Yet the exact same chemistry used in my phone, which is a much lower peak pulse load, and a low C rate that is charged daily is 2 years old. The relationship between current and cycle life is huge. So the difference a 120 to 180Ah pack makes to longevity and reliability of all lithium is huge. It makes looking at cycle life figures almost meaningless. Which is why most manufacturers quote cycle life based on fractional C rating or at best 1C.

The next thing is charge volts, depth of discharge and frequency. With a 40Ah pack you will get just 25Ah from it if de-rated in capacity to avoid the effect of trying to get peak Ah from the cells. Tesla charges to 4.1 volts, and only discharges by 70% DOD or 30% State of charge minimum. Thats how they unlock an extra bunch of range in times of floods or tornados as seen on the news. This is why:
Please spend a little time looking at this page, it is accurate and informative, and relates to lithium ion chemistry that you are talking about.
https://batteryuniversity.com/learn/art ... _batteries


There are many things that shorten the service life of lithium batteries.

That capacity rating you quoted for the cells I linked to is wrong. That is their WARRANTY. It is expecting them to be used with a small pack, and with a typical BMS and those things universally screw up LiFePO4 cells.

The actual well tested and proven cycle life of LiFePO4 cells is:
1500 cycles at 100% DOD to an end point of 80% original rated capacity.
2000 cycles at 80% DOD to an end point of 80% original cap.
6000+ cycles to 60% DOD etc...
I have the full detailed test and spec sheet here should you wish a copy?

Those specs are when charged to a torturous 3.650V. We generally use a very controlled charger, accurate to 3mV, that has a 1A balance control, and is highly programmable. And we charge to a lower 3.550V per cell. That costs around 1% of capacity on a LiFePO4 cell. That alone, means more cycles that the specification.

And because we use typically 120Ah to 180Ah we are loading the cells very lightly, at around 1.2C peak. And a fraction of 1C much of the day (week!) And because a large pack gives us an enormous range, then we seldom need to charge more than say weekly in winter, or maybe twice a week in summer that multi thousand cycle life is decades. And no more range anxiety, the pack runs down slower than you do.
And because that end of life point is 80% of 60 or so miles, its not really end of life at all. Because even at 10,000 cycles we still have more range than lead! Or a 40Ah new pack. In other words the battery will outlast at least 2 powerchairs. And we dont burn. And they are cheap even doing a DIY setup with 5x the range!

Dont get me wrong. A small firey lithium pack that is easily unclipped and removed in the event of fire does make sense on a travel chair, or a small scooter. Or even a hover board balancing thing like all those that were catching fire 2 Xmases ago. And thats exactly where you do find them.

[Scooterman]

It is seriously NOT 'idiot proof' and a user can quite easily destroy the pack with some simple mistakes....



ex-Gooserider

This is true. But if you only use a pl8 for charging a powerchair you only need ONE preset and can "hide" all the others. At least I think you only need ONE preset? Ooh I'm getting worried now

I'm in need of reassurance

[Scooterman]

Some light bedtime reading for you. viewtopic.php?f=2&t=1813

HA HA!

[Swan T.W.]

What about this charger? https://www.ebikes.ca/documents/Grin_Sa ... .3_WEB.pdf

[JoeC]

I have the full detailed test and spec sheet here should you wish a copy?

I'd love to see it, thank you!

You do make an excellent case for why my original proposal of derating a 50AH pack to 40AH wasn't conservative enough. Still, I don't know if I see a way to get lithium into a commercial chair unless it can be flat-out cheaper than a set of gel cells. The cells used in cars and stationary storage are heading toward $100 for 2kWh of capacity. Is that true for LiFePO4? If it's not, it doesn't matter that you could ammortize the cost over the life of two chairs. If it goes to twice the price of car cells then maybe it could have less safety packaging and operate at a higher C rate and save cost there, but I think that further cost reduction is necessary before any existing company will pick up the project.

[Burgerman]

MK gel, in UK are £250 each. So £500
That gives you a usable peukert reduced 40Ah before the chair grinds to a halt.
Thats £12.50 per Ah.
They last me approx 9 months, so thats repeated every 9 to 12 months, while you suffer limited range, and range anxiety. And super slow charge times. And double the weight of LiFePO4 of 4.5 times the Ah.

I can order a 3C 180Ah pack (16x LFP 90Ah cells) from china, for £800 today INCLUDING DELIVERY in DIY quantities of just 16 cells. Thats £4.44 per Ah. So already ONE THIRD of the price of lead. And no peukert. And so we get an expected 10 year absolute minimum life and massive range. I already have a LiFePO4 pack in a chair here thats 8 years old, and it still measures withing 7 percent the same Ah as when new. If we ignore this and cost it only over 3 years (which is ridiculous as it will easily do 10+) Then thats as per the following:

LEAD MK replaced every 9 months, over 3 years, needs 4 sets. Thats £2000 for an actual 40Ah available.
LITHIUM of 4.5 times the capacity (same storage as 9 MK gel batteries) is £800 for an actual 180Ah available!
And it will outlast 2 chairs and still have 4 times the range left over...

So over 1 year its £300 more expensive for 4.5 times the range, and a lifespan measured in decades.
If you dont use the chair much, you could change the lead battery every 2 years. Now the lithium is ALREADY £200 cheaper over 2 years. And still has the weight, range, 10 year service life, safety, fast charge capability, no service or work required, etc.

So I dont understand why you think its more expensive. I simply cannot afford lead bricks today!!!

[Burgerman]

Cant find it right now. But heres the one for the 90Ah cells mentioned above.
16 of these fits into the space of a set of grp 24 batteries. With some free space at one end.

These are ONE THIRD the cost per Ah of MK gel today at retail prices. After delivery cost from china. Which is 1/3rd the cost.

So allows 180Ah, and 3C peak, so 540A capable. This means a long easy life as the peak of 200 to 240A and the low average discharge rate doesn't reduce service life. Remember that with a pack this big we can stay in the long life sweet spot, low current per Ah, and between 30 and 85 percent charged. So they get an easy life.

[JoeC]

It's more expensive because 800 is more than 500, and once the chair leaves the factory then the wheelchair manufacturer isn't who's paying. If they had to keep supplying batteries, then what you're saying makes sense, but they don't.

So, how do we get the manufacturer from pushing this cost off to the user? The answer is probably a massive political campaign to change that financial structure, but I'm an engineer so I'm looking at engineering solutions for the existing financial structure.

[Burgerman]

But those are retail prices. And the lithium includes around 30% shipping.

So the reality if ordered in bulk, is that they are the same price if you compare 180Ah lithium to 70 gel, (40Ah usable)...

Compare Ah per Ah, INCLUDING extortionate cost of shipping 8 cells from china, £4.44 per Ah for lithium to £12.50 per Ah for lead. So ignoring longevity its already 3x cheaper like for like! *** But if you do that then you lose all the longevity, and range advantages. Which is the whole point! So you SHOULD fit more Ah.

However, if you compare a 6 gallon lead fuel tank, to a 40 gallon lithium tank, yes they are the SAME price if you ignore the shipping from china in small quantities.

[JoeC]

Ok, so it looks like the LiFePO4 chemistry might just be cheap enough. I think that a chair manufacturer would select less than 180AH and aim for the battery to wear out at some specific point, maybe five years of "normal" use, but it looks like if a manufacturer wanted to do it then they wouldn't be totally blocked from the chemistry. That's good.

So, what else? Updating the controller to have built-in BMS would be a large development cost, but probably a one-off expense with a lot of testing to show it's been done right. I think the largest remaining hurdle would be in shipping. What's it like to ship those batteries from China? Do they go by air?

[Burgerman]

They can. But sea is cheaper. Usually by a known carrier.

BMS...
BMS is 2 halves.
1. Charging (which they are dismal at)
2. Protection from over discharge Amps, too low volts, too low per cell, too high per cel, etc.

2. first... The most complex and expensive protection half, isn't actually wanted. The battery, when choosen correctly, and large enough, when used with a complex programmable controller, can and does already do this stuff, in a far more controlled and proportional way.

Other than individual cell level monitoring. Which again is only really needed if you buy a pack, thats not matched well, and its small enough that you will be discharging anywhere near this level. So the ONLY thing a controller must do that they do not already do is monitor the lowest cell voltage, both ON LOAD and at rest. It must know that a 2.9V cell may drop to say 2.3V under heavy load, and ignore that, but reduce power proportionally if it approaches say 2V lowest cell under load so that power doesn't cut out, as it does with a dumb BMS, so it allows you to remain in control. Again this is less of an issue, if the pack is bigger.

I can honestly say in 8 years, with no monitoring, and a big pack, I never came close to getting low enough to matter. The smaller the pack the more you lose out on the lithium benefits.

We will go back to 1.
The chargers we use have a 1A balance circuit. Are 40A so can charge even a large pack in reasonable time. And unlike lead a full charge isnt essential. So we charge easily at 40A or at 16 to 20 miles range added per hour of charge. But more importantly, we can choose a lot of charge parameters. Inc charge voltage etc. And the thing gets balanced accurately, and the charger is throttled proportionally, so that the balance current isnt exceeded. This means controlled proportional balancing and no cell exceeds the choosen voltage.

A BMS isnt that... They typically have 50mA to 100mA balance. They cannot control the dumb charger, so they charge at a lower rate, and rely on letting the highest cell go above the choosen point to anything up to 3.8V and then cutting off the charger, while the high cell drops to 3.5V then rinse and repeat... So you ruin the cells, takes hours to days to fully balance, and you never know when thats happened. So ideally a better charge system built in like a black box pre programmed PL8...

[Burgerman]

Or just do it externally, with one cable, 40A and balance all in one.

http://www.wheelchairdriver.com/BM3-con ... le-40A.jpg

And a charger.

Then we can see whats happening and spot any issues should there be any if you wanted on screen.

This is all any of us do here, no BMS at all, and we have no issues.

Testing a couple of lithium cells here. For capacity. Messy bench.

http://www.wheelchairdriver.com/BM3-con ... arging.jpg

Charging 45V 13S pack

http://www.wheelchairdriver.com/BM3-con ... arging.gif

[shirley_hkg]

And here . . . . .

[Scooterman]

[Burgerman]

Thats to 3V if you look at the rebounded voltage. Full capacity is down to 2.5v which isnt very good for them to do unnecessarily. But theres another 3 to 5Ah down there.

[Scooterman]

Thats to 3V if you look at the rebounded voltage. Full capacity is down to 2.5v which isnt very good for them to do unnecessarily. But theres another 3 to 5Ah down there.

Thank you BM. I was half joking because I understand the basics of lithium, but the PL8 charge/discharge voltage, current,etc settings still confuse me.

I never had much success with those tiny 600mAh cells I bought. I don't think I ever got to a completed balance charged position, even after hours and hours. It sort of knocked my confidence a bit about lithium. But I trust what you and Shirley say, and if you both say the lf90s are good that good enough for me.

"When you don't know something, defer to wiser heads".

[shirley_hkg]

Thats to 3V if you look at the rebounded voltage. .

That were under load. Rebounced to 3.1V