Announcement

Collapse
No announcement yet.

Cautionary tale for custom battery builders

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Cautionary tale for custom battery builders

    Hello all,

    I stumbled across a terrifying video of an e-bike battery exploding into flames in an elevator. There was a stroller, and three other people on that elevator. No idea as to make, or configuration of the ebike or battery.

    I mean, Christ on a cracker! Look at this older video.



    It really impressed upon me yet again how incredibly dangerous these batteries can be. There is a ridiculous amount of stored energy in there, and we MUST be extremely careful in how we design, build, and ensure safety at all stages when making a custom battery. I am a firm believer in double wrapping cells, and I used a plastic separator tray in my large battery build, so that batteries are not in direct contact with each other. I was going to double wrap, and directly stack batteries in my 48V pack build, but that video convinced me to expand the case design and use those plastic separators again.

    The video also reminded me of why I DO NOT simply hunt for the cheapest battery. You really need to be careful when selecting a battery for your application, and be positive of what kinds of duty cycles you will expose it to.

    Jose
    Last edited by DaHose; 05-11-2021, 09:00 AM.

    #2
    Something else to remember

    Not all homeowner's or renter's insurance will cover fires from electric bike batteries

    Comment


      #3
      It is interesting that packs don't seem to explode while being used. I have seen more than a few videos kinda like that one where the pack was recently in use but then has sat for some period of time often perfectly stationary and they suddenly go up.

      I bet what happens to a lot of these that blow up when they are not currently under heavy use or charging is vibration let things move enough to short out a single cell. That one cell heats up to the point it goes and that heat could be enough to set off another or at least cause a short that sets off another.

      That sort of thing is where I feel good about the extra money I paid for a pack like the Luna Wolf. Fully potted so nothing is moving at all so greatly reduced risk of going boom. It can't be repaired either but big boom with collateral damage vs a pack that maybe just doesn't work seems like a fair trade. Both are fairly rare possibilities but if you had to choose one I would choose the one with no flames. I should look for videos and or pics of a Bosch e bike battery apart, my guess is one way or another those are pretty solidly built. You can't say that about many of the packs you see on E bikes, many of them are either just soft packs or naked packs barely hot glued together stuck loosely into a plastic case.

      I am also told (not by Luna) that the potting is also very good for thermal conduction. Air isn't a good conductor so as long as they used the proper type of potting material not only do the cells stay in place they also can shed heat more efficiently. Having an air gap between cells is good if you have some sort forced air flow around the cells (or a liquid cooling circuit like Tesla) but very bad if you then basically seal them up in a wrap or case. Thermal wasn't a huge concern when we were charging and discharging on the miliamp scale but now when we are taking a full amp or more per cell its a concern.

      Comment


        #4
        The RC community went to using fire-resistant charge boxes eons ago. It's crazy that people are not learning from those lessons. I saw a guy who's garage caught fire from a large capacity RC battery charging on his bench. Took out his Porsche Carrera too.

        I think a real problem is that we get so used to almost all electronics having LiON batteries nowadays, that we let our guard down. All other electronic, and even cordless tool batteries are under WAY less load at all times, than these ebike packs. People just drain/charge at high rates, without thinking about consequences. We all know to monitor motor, and battery temps., but that's because we actually frequent forums like this one. Your average person just has no idea what is going on in their batteries.

        Jose

        Comment


          #5
          The different chemistries have different thermal runaway potentials. Battery construction and BMS are important potential failure nodes.

          Aside from the less mass efficient lithium iron most all have flammable electrolytes.

          Say you have a 1KW space heater... these batteries will put out that much heat for from many minutes to us much as an hour - think of it all leaving in only a couple of minutes... they reach a point of thermal runaway where the heat they create increases the battery/cell temperature which increases how much heat the battery/cell are making... and so on

          Electrodes get dendritic growth/creepage (they all do), electrolyte gets wonky, high current paths in cells happen...

          It seems to be most likely in charging but it's hard to say really... at least it seems there's no way high discharge can't not be a factor... There's a lot of videos of fires or aftermaths... don't know what the data says if there's enough to draw conclusions - videos are anecdotes not data

          Just know they deserve respect. Like DaHose when fussing with LiPO's (high on the dangerous chemistry list for several reasons) I only charged them, transported them, stored them in fireproof boxes.

          There are different chemistries though... some have more potential, some more likelihood for catastrophic failure

          The lithium manganese cobalt, etc. we typically use in electric bikes are getting well tweaked for our kind use - moderate-high for power delivery, for life-cycles, good max power, for safety, cost, weight, etc.

          Comment


            #6
            I tend to think that a vast majority of people will buy ebikes modded for high speed/power, but don't think about the impact to the battery stacks. They will just blindly run at full throttle, at all times, and not monitor their systems at all. Would not be surprised at all to find that is a major contributing factor to many catastrophic failures.

            Jose

            Comment


              #7
              I'm sure that is a factor but I think the vibration letting things shift and short is also got to be a factor. You look at how poorly secured many packs are the cells can rattle around and shift so that the connection strips also start to twist and flex. I have not looked close but don't most 18650 cells have the neg part of the can still roll over the top a bit so its awful close to the pos cap? You look at how cheap and poor the case insulation is often done and the fact that you have a series parallel stack going it doesn't take much movement to make things touch that should not be touching. Even if there is some wrap on the cell how much vibration does it take to rub through it?

              I suppose you also look at how many packs are not what they say they are so even if the person is kinda aware that they should have a pack that can do say 30 amps are they actually getting that? They may be getting no name cells with a no name BMS and some of the cells could be full of sand. Motor pulls all it can get and nothing in the pack to stop it so things get hot.

              Spending money on a good battery just seems so hard for so many people. I'm currently helping someone do some upgrades and the guy has maybe $4k into the bike and extras now and keeps sending me links to $250 battery packs that claim they are better than my $600 Wolves. Maybe Luna is making a couple bucks off the sale of the Wolves but there isn't no $350 in profit there. You can barely buy the raw cells they claim to have for the $250 so that tells you something has to be wrong right there.

              Comment


                #8
                Originally posted by DaHose View Post
                I tend to think that a vast majority of people will buy ebikes modded for high speed/power, but don't think about the impact to the battery stacks. They will just blindly run at full throttle, at all times, and not monitor their systems at all. Would not be surprised at all to find that is a major contributing factor to many catastrophic failures.

                Jose
                At the risk of sounding argumentative pretty much none of the riders I know are interested in high speed or power... none of the fellows I ride with have anything more than "stock" BBSxx's and some are running rear geared hubs too. The guys I ride with fall into two camps... those that want to ride the multi-use paths around town, going ~15mph and getting ranges of >30-50mi on 10-14Ah batteries... the other camp takes their bikes out to the backcountry and ride even slower for the most part... some of us do both...

                Frankly of all the people I know with electric bikes I likely run the hardest - my normal riding is just like the two camps above... unlike those folks I do occasionally run hard >20mph for short bursts and when off-road I'll climb steeper harder stuff than them

                I inherited a monster 14s8p battery and it opened up some higher power/speed options but even when I'm going solo around town with that monster my speed only goes up to around 20mph and I have yet to run it more than 750W continuously

                Most of my riding friends are in their 50's and 60's

                I agree that most people don't pay as much attention to their batteries as they likely should. I don't think a lot of users understand how large a portion of their electric bike cost is the battery. The folks I know have a really good idea of the costs and most at least understand not letting a discharged battery sit around or letting a battery plugged into a charger for months on end and like mentioned they're not flogging them to begin with...

                Comment


                  #9
                  When I started hanging around the forums to learn all I could, I repeatedly saw those with experience point out that your battery will likely cost as much as the motor/bike combined. That is a big part of what prompted me to build my own. I wanted to know for sure that I had high quality cells, and it fed my project junkie habit to have something else to build. Cheaping out on a battery is something that anyone with experience won't do. It's unavoidable to have factory defective batteries and all, but some of us here have even melted our connectors on the batteries, but been protected by a good BMS. That kind of thing is what makes me think it's the uninformed and "careless" that are really ending up with the catastrophic failures shown in the video.

                  Your experience of how people around you ride (AZguy), now makes me wonder how many of those catastrophic failures are from people riding like you vs. e-moto type riders who just want high speed throttle drive. Perhaps what is going on is not really statistically significant? Since there are a lot of ebikes out there now, maybe we are just very quickly/publicly hearing about all the outliers who aren't nice to their equipment and "causing" failure through their mis-use or abuse. That is the issue you see with negative feedback of vendors. For every 1000 satisfied people there might be 1 dis-satisfied, but the angry one is the most likely to leave feedback that is read. So if you have 99 good reviews, and 1 failure report. It's 1% failures. But if it's really 999 satisfied customers, and one dis-satisfied that you hear about, it's actually a .01% failure, which is an interesting thought. Consumer electronics have failure rates as high as 5%, while automotive standards say .002% is acceptable. So 10 failures per 1000 batteries is 1%, which is WAY under commercial electronics, yet WAY over automotive standards.

                  I suppose this hypothesizing would benefit from real world numbers of how many catastrophic ebike battery failures are reported, vs. how many are actually sold.

                  Jose

                  Comment


                    #10
                    I think you make a lot of very good points. The interwebs tend to magnify the extremes and often the negative extremes.

                    I'd bet most battery failures are people that don't give them much though and just ride their bikes without flogging much simply because it's my perception that's by far the majority of owners... we see a bunch of builders here

                    Of my friends with bikes (electric or push) almost none of them do much wrenching and what little they do is often over at my workshop! After all I've got the stand, the tools and.... errrr.... the elbows =]

                    Some wouldn't even touch them just to lube a chain let alone fuss with things like derailleurs. I had one buddy that would just leave his electric bikes parked for months on end (beach house) and he just couldn't get it through his head that you shouldn't just leave them plugged into the charger all the time they sit unused - it's a guaranteed way to radically shorten the battery life but I don't think he would get it until he destroyed a battery... he just wanted charged bikes when he was out at the beach house and didn't want to think about it...

                    Comment


                      #11
                      Originally posted by 73Eldo View Post
                      It is interesting that packs don't seem to explode while being used.
                      Check this one out, caught fire while he was on it.

                      Aftermarket battery, likely in parallel with a completely different battery, oh and it didn't even have a bms. Guy must have a deathwish.

                      Comment


                        #12
                        Originally posted by paxtana View Post

                        Check this one out, caught fire while he was on it.

                        Aftermarket battery, likely in parallel with a completely different battery, oh and it didn't even have a bms. Guy must have a deathwish.
                        Ignorance is bliss...... errrrrr... mebbe not... Click image for larger version

Name:	headscratch.gif
Views:	162
Size:	418 Bytes
ID:	128064

                        Comment


                          #13
                          Ya parallel miss matched seems like an even worse idea that parallel in general.

                          With a typical bike pack which has a '3 wire' BMS which is the separate charge and discharge port as long as the BMS is happy and 'ON' if you parallel the discharge ports the packs are pretty much direct in parallel and can charge and discharge each other all they want right? On the discharge port most BMS's just basically have a On/Off switch and that's it? Most don't even have any active current limiting, the published current limit is more of a physical construction limit? As long as the pack voltage is above 0% its ON? So parallel you really get no protection or balancing or anything other than a low volt cut off?

                          Now if you go to a 2 wire BMS paralleling isn't quite as bad because if the BMS senses higher voltage coming in it goes into charge mode and will do balance and over charge protection ? One pack would in theory safely charge the other till the balance? So if you really wanted to parallel getting packs with 2 wire BMS would be safer not to mention would allow you to charge with them still connected?

                          How common is it to not have a BMS or to have a BMS fail in an unsafe way? My only failure so far was it just got over protective. I suppose since they build fake cells that someone out there must build fake BMS's.

                          Comment


                            #14
                            I seen a couple faulty bms fail in unsafe ways, one where the part that drains down cell banks did so unevenly and caused cell imbalance, and another where that same component got too hot. But yeah the vast majority if they fail would fail in a way that is overprotective as you described

                            Comment


                              #15
                              I was a little unsure about using a BMS when building my batteries, as some people kept saying they preferred to use an external balancing board. Now I think that really only applies to RC batteries for applications where weight matters. With what I have since learned, there is NO WAY I would build an e-bike battery pack without a BMS.

                              Jose

                              Comment

                              Working...
                              X