mpetry912 

So here's my current working theory on battery fires. I get the "cheap" battery with no BMS or thermal cutout, OK. But:

Batteries have VOLTAGE and CURRENT. Most ebike batteries are 36, 48, or 52 volts. But as the battery is drained by riding, the voltage drops, and the controller / BMS has to supply more current to keep a constant level of power to the motor. That increases the battery temperature.

Make sense so far, right ?

So at very low voltage levels, if the battery is drained to say 10 % and then you put it on the charger and leave it powered on, that could cause a thermal runaway resulting in a fire.

I seldom drain my bike's battery below 20% and I don't leave it charging unattended. The charger gets noticeably warm but I keep an eye on it.

if you leave a battery charging unattended and plugged into a power strip with crappy / old house wiring, or overloaded outlets, then all bets are off.

"UL Approved" won't solve that problem. I'm thinking too of that auto transporter that sank off portugal last year taking 100s of electric Porsches and mercedes with it. What do you suppose happened there ?

I think somebody got in a car to listen to the radio or turn on the heated seats, got out and left the power on.

anyway that's my theory ! I'm sure I'll get "flamed" in the responses.

/markp

veganbikes

You have a quality bicycle that is unlikely to have those issues. UL certifications will certainly help and good BMS' will definitely help. If you do have a thermal runaway and the BMS can stop it you are extremely less likely to have a fire. Plus if the battery is rated by UL or another similar company you can know that battery was tested and passed some decent tests. No it won't solve all the issues but having a product rated to certain safety standards does make a huge difference. I would rather that then just unregulated stuff as we are seeing with the fires.

If you have crappy wiring than that could cause a fire but it might not be the battery or might get the battery in the process but it might not be the battery causing the issue. I personally leave my bikes unattended all the time though I do take some precautions to make sure I won't have issues. My charger does warm up a bit and in one case is the same charger as you (but on a Como not a Vado) but I keep it away from significantly flammable items. Having a Bosch equipped bike and a Specialized Brose equipped bike have lead me to not worry because I know they have put a lot of time and money and R+D to making a safer battery and of all the batteries we have had issues with which is about 5 in 8 years none of them have been from the Big S or Bosch.

I am not saying one shouldn't be careful as you should I just trust those brands from years of experience with them and personally am confident I am very unlikely to have issues.

In terms of the ship fire, I don't recall there being an knowledge of how the fire got started and I don't know how many of those vehicles were EVs and I know there was a bunch of non-EVs and some older vehicles as well. However in a big fire yes those batteries could easily be a problem but maybe it wasn't the battery that started the fire there were gas powered vehicles on board.

I hope you aren't leaving your bike with that lock though. Forget the fires, that thing is nice and should get one better protection.

Smaug1

(Former UL engineer here, worked there for 21 years and part of that on battery chargers and lithium power tools. (electronics engineering background) Now I work for a tool company that builds lithium power tools and I'm the regulatory engineer for them.)

UL Listing on battery packs and chargers would mitigate almost all fire concerns. Yes, there are a lot of tests that the products would have to pass: both normal and abnormal. (forseeable misuse) Manuals probably say to not charge unattended, but that is unrealistic. No one (or very few people, anyway!) is going to sit there and watch their bike or battery charge for 4-8 hours.

A UL Listed battery pack would have to use certified cells, which have CIDs built-in. (Current Interrupt Device) If the current gets to fault levels, this CID would pop and open the cell. Then, the BMS (battery management system) would be tested. Ideally it would be tested in combinations: battery+bike and battery+charger. There would be component faults in the BMS. With the charger, a cell imbalance is introduced by discharging one cell or parallel cluster of cells, then putting them back into the battery circuit, then charging the pack. Since chargers usually only measure the whole pack voltage, this is a tough test, because it means the rest of the cells in the pack would be overcharged for the pack to get to the cut-out voltage. There is the shorted cell test, too. Of course shorting the discharge contacts of the pack is a forseeable misuse, and also overloading of the charger, due to a faulty or tired battery pack.

As for temperatures, a warm charger is nothing to worry about. The external temperatures are only limited to the extent that they shouldn't burn a person putting his hand on it. However, remember that if the charger is 50°C (122°F) externally, how hot is the transformer inside? How about the PCB that supports the current-carrying traces. It's a LOT hotter inside. UL measures internal temperatures, usually 15-30 locations to make sure everything is running within limits for each component. Electrolytic capacitors sometimes have pretty low temperature limits. PCBs are usually rated at least 105°C.

After testing the battery & charger, UL will control the safety-critical elements of the construction, so that the manufacturer cannot just change parts because they found something cheaper. UL then conducts un-announced inspections quarterly to make sure the manufacturers are still building the UL Marked products in the same way, year after year. The whole process is onerous and expensive, but this oversight is often needed to keep people safe. Remember when unregulated "hoverboards" were burning houses down? The government asked UL to write a standard for them and then required that incoming hoverboards were UL Listed. It's only a matter of time before the same happens for eBikes. Some manufacturers are already doing this voluntarily, having already been bitten in a few lawsuits.

I wrote this all out to point out that UL Listing can often be correlated with a certain level of quality. It's not causal, but I think you get the idea. Someone could still build a PoS product that is UL Listed, but at least it would be a safe PoS, hehehe. It's one of those things that doesn't matter until it's wanted, like a seatbelt.

I just ordered a multi-voltage charger from aliexpress. As far as I can tell, it's not UL Listed, but I'll have to wait and see which regulatory marks are on it. CE (for Europe) is self-declared by the manufacturer, so that doesn't mean anything on its own. It is just the manufacturer declaring he did the right thing. If asked by an inspector or government, he has to prove it, or else there's a big financial penalty. I'm gonna have to keep an eye on it. (I bought it so I could use it to charge a 48 V pack to a lower voltage, so that it is only really about 80% charged)

One thing I've noticed can be a fire hazard that is not often thought about are high impedance connections. An everyday example: Feel with your hand the temperature of the plug/cord junction on your toaster. Those often run hotter than they should. The impedance of the connection is too high, due to cheap spot welds, and power is lost there, turned into heat. If they're running warm, I cut those plugs off and fit a (UL Listed) mechanical clamping type plug from Home Depot. Notice how much cooler the plug/cord area runs after that. The same thing can happen with eBike charging connections. It's a decent amount of power being transferred there. On my heybike Ranger, it had a metal four pin barrel/sleeve connector, but he sleeves loosened up after a couple of charges and it was always warm there afterwards. I didn't like that at all. The two contact concentric connectors are usually better, in that regard, but they're getting to their power limit when charged at 60 V and 3 or more amps. UL has temperature limits on the plug prongs during their testing, but I suspect many products barely pass. Add a bit of manufacturing tolerance and ...

mpetry912 

Thanks. it's a nice bike, I've upgraded it with DI2 shifting and built different wheels with Chris King hubs.

I'm fortunate to live in a place where there isn't much crime. That lock is fine for a coffee shop lockup. In downtown Seattle - would need something more substantial

/markp

Smaug1

I think the most likely cause of fires for batteries is the charger or BMS not being smart enough to detect fault conditions.

My company's batteries/chargers monitor temperature and current and will cut power if either one exceeds limits. They won't start charging if pack voltage is too low. (indicating a bad cell, which would lead to overcharging I mentioned above)
In eBike batteries, I don't know what is done in this regard. With a two pin connection going into the pack, there's no room for that kind of sensing, and we would hope there is some well-designed smart circuitry built into the BMS in the pack. The less expensive the battery pack and with the fewer certifications, the less I'm sure of it.

Another one that causes fires would be using the wrong charger. Let's say you have two eBikes; one is 36 V and the other is 52 V. The 52 V battery would be charging at around 60 V. Mix that up and plug it into a cheap 36 V eBike battery without good BMS built in and it's a recipe for a fire. (lithium batteries will light off if charged at too high of a voltage. It's 4.2 V per cell or parallel cluster, IIRC)

downtube42

Having spent most of my career working in industries and companies that seek certifications (self or otherwise), I'd add that products are just one middle manager or low level executive away from being fraudulently certified. From mid executive up, they just want to hear it's all good. From front line manager down, they just execute based on pressure from middle management. In between, that's a **** show.

unterhausen

I expect a lot of the fires are from batteries that are basically fraudulent products. If you randomly buy cells on alibaba, you will eventually get some that have lower capacity cells inside. But it doesn't even have to be quite that bad of a scam before the cells will give you problems. Sony had laptop fires, it's not like it's really something you can just decide to make without some technical expertise. And a lot of people don't have a good charger either, so in combination with borderline cells that's a fire waiting to happen. I feel fairly confident that my wife's Shimano battery is safe, but it's a little bit of a worry.

7up

Safest battery chemistry is “LIFEPO4” period. Stop buying garbage batteries.

CliffordK

My power tool battery chargers won't start charging when the battery is hot. A few minutes ago I just let the whole thing cool down for an hour or two before putting it on the charger because I didn't need it until tomorrow.

Perhaps one should also consider charging speed. If one doesn't need the battery to be charged for 10 hours, then there is no reason to charge in an hour or two. Just start it up overnight and let it trickle charge.

tds101 

Can you personally recommend ebikes with LIFEPO4 batteries?

unterhausen

I didn't realize there were as many options for LiFePo4 ebike packs. I guess the energy density isn't quite as good as Liion, but that probably isn't much of an issue.

tds101 

I definitely see tons of options for replacement battery packs, but are there actual ebikes that have them? Like, without a seperate purchase? Almost all use 18650 or 21700 cells nowadays. When Googled I see 21700 LiFePo4 cells as well. So confusing.

2old

I haven't heard that any of the "majors" are using LiFePO4, probably since (I've heard) the batteries are a lot heavier and there's a "weight war" between them now.

Smaug1

I feel like you don't know what you're talking about. LiFePo4 is SAFER than LiIon, but with a lot less energy density. It's not widely available in eBike battery packs. I believe its energy density is between LiIon and NiMH.

When I was heavy into R/C airplanes, LiIon was considered too heavy and weak; we used pouch-type LiPo (lithium ion polymer) which is even MORE sensitive. I'll be surprised if that segment hasn't moved onto super caps soon.

2old

From endless sphere comments, it seems like some of the most ardent supporters of LiPo have moved away from it to LiIon. However, nobody mentions LiFePO4.

unterhausen

I have some lifepo4 batteries that I was going to set up for RC airplanes. Like a lot of my projects, it never got finished. Walmart discounted some battery packs for tools that used that chemistry so I got them cheap. My RC charger will charge them, so it was popular enough for charger manufacturers to notice. There was a U.S. company that was promoting it heavily, they even had a drag racing team. Lipo discharge rates got a lot better since then and I suspect Liion did too. Discharge rate is pretty important for an RC plane, and the higher the discharge rate the faster you can safely charge them.

I think the issues with Liion packs probably follow LiFePo4 as well. If it's not from a trusted supplier, it's going to vent with flame under similar conditions. It wouldn't surprise me if some LiFePo4 packs have cheap Liion cells in them, it's the nature of the industry that the low-end is often fraudulent.

m@Robertson

I would love it if LiFePO4 was a viable solution for an ebike. I use it for my portable power station and my home system, which have 2 kwh and 9 kwh of storage, respectively. The energy density is just too low to be practical. I've carried my 2kwh unit and its like 65 lbs. Each of my three 3kwh home units are literally 99 lbs. Scale that down to ebike size and I bet we are adding 10 lbs to a bike. My 35ah 21700 pack on my most recent cargo bike build... I don't even want to think about how heavy that would be in LiFePO4.

But boy I sure wish it was workable cuz I would really like to have all those extra charge cycles, and the tolerance for high state of charge.

Considering the solid state cells that are in pilot production now, and how long it took 21700's to go from this same phase to commonality as they finally are now, I bet it will be at least another five years before we are able to look past the Li-NMC 21700 cell to whatever is next.

Pop N Wood

There is little that a BMS can do if a battery cell shorts out internally. The current starts flowing inside the cell, it gets hot, causes the cell next to it to overheat and in turn break down/short out, creating more heat and eventually a fire. Hence the moniker "thermal runaway". Once a cell shorts out, it is going to keep arcing until is expends all the energy stored inside of it. All one can do at that point is cool the battery pack to keep the burning cell from taking other cells with it. Hence the reason airline attendants are trained to drench an overheating lap top with water, cool it down and stop the thermal runaway

It can happen to even the best of batteries, really doesn't matter who makes them. A quality battery will have a much smaller probability of it happening. Manufacturing tolerances, impurities, poor manufacturing practices, and any sort of physical damage can all cause a cell to short out internally.

To the OP the BMS can't "increase the current" to supply more current as the battery is depleted. Whatever juice the battery can supply is a matter of physics. If a cell is depleted or dead, the internal resistance increases limiting what it can put out.

mpetry912 

I get the part about the internal shorting of an individual cell.

OK, if that's true, then let's consider a battery that is fully charged - to say 48v, 100%.

Now you ride it and get it down to 50 % remaining. The voltage is something less than 48 volts. With me so far ?

So new explain why, even at 50 % indicated battery capacty - or 25% for that matter - the bike is really no slower than it was at 100%

if the current draw from the battery does not change with regard to voltage, then you have either violated ohm's law or the laws of physics.

explain !

/markp

cat0020

Micah put out a good video yesterday.
After converting dozen bicycles into ebikes and owning/riding few dozens of ebikes for nearly 10 years, I agree with the video:

Calsun

Lithium batteries are all prone to thermal runaway. When it happens it is not the fire as such but where the device happens to be. If it is while in an aircraft it is more serious than in an automobile.

With electronic circuit boards the more individual chips on a board the greater the probability of one failing. With cars like the Tesla that have a great many individual cells the odd of a fire increases dramatically and the manufacturer adds circuitry to remove individual failed cells and this works most of the time.

Fire codes mandate smoke and fire detectors in houses and they should be added to garages where an e-bike or electric car is kept.

2old

Might not be long before insurance companies won't insure individuals who have Li batteries in their houses/garages or charge an inordinate premium. That could be a dilemma for some (me).

Pop N Wood

My comment was directed toward what the individual cells can do. The voltage and current the cell is capable of putting out depends on the state of the cell itself. Any downstream electronics won't change that.

To answer your question, the motor controller sits downstream of the battery. If controls the current applied to the motor. It is only going to allow a certain amount of current into the motor. If the battery has excess capacity, then you may not see a drop in performance with the state of charge. When the battery gets weak enough that it can no longer supply the desired current flow, you will see a drop in performance.

FWIW I feel a drastic difference in performance when charged to full volts vs. even 80 or 90% charge. I have a 50 amp battery while the controller on my bafang limits the motor current to 30 amps. So if the current stays constant, the power into the motor drops as the battery voltage decreases. Your bike has more limiters, it is my understanding that bikes such as yours are limited to some power level. My guess, and it is only a guess, is the motor controller varies the current limit with battery voltage to keep the power in the same. Since your bike almost certainly isn't allowing 1800 watts into the motor, you will need to really discharge the battery to a low state before you will see any drop off in performance.

mpetry912 

yes, I think we agree. thanks for the explaner. clarified my thinking some.

/markp

Leisesturm

When e-batteries catch fire while enroute somewhere in the cargo hold or container of a plane or ship ... are they hooked up to chargers? I imagine battery fires are caused by the failure of insulating components meant to keep high potential areas away from one another. I try not to forget, but every now and then I'll remember a 4x18650 battery pack left on the charger for two days. Good thing ALL my battery packs are mid-tier if not top of the line. My mid-drive battery 52V x 17ah cost over $500. I don't worry about it catching fire. I had to get off Endless Sphere because of all the 'bomb disposal' equipment posters were insisting I needed to charge my battery.