Self-heating lithium LiFePO4 batteries tested - FAIL!

[@KempOutside]

Excellent video!

[@robfodder5575]

​ @roamerbatteries Its odd because there are hints this is a Fogstar battery over social media, yet they showed a pic of the sensor between two cells where film is not between (heating is between the cells, just not all cells from the pictures they've shared in response to your video). Most EEs deal with control algorithms and managing temperature for many products and from that standpoint the uneven heating seems bizarre, trivially simple issue to resolve. What I'm getting at is instead of rubbishing the competition, you'd be better designing a solution that works and highlighting what you do differently, especially as this is such a cheap change and mostly in software too! I suspect, as someone else has suggested on social media, this is because you import a cold climate battery and thus the aforementioned would harpoon your own sales.

[@roamerbatteries]

While that does sound like an improvement on other models, our concerns go further than simply where heat film is placed. The use of large prismatic cells (up to 7cm thick and weighing 5kg) makes this incredibly difficult to get right. Better temp placement and wrapping all cells would be a great start but really I think this is only practical if using small cylindrical cells. To be clear, we have not made any suggestion of specific brands and don’t intend to. This is not a name and shame exercise, we just want to highlight our concerns with the way this is implemented. Hopefully this is taken on board and improvements are made. At the very least, I would hope that additional testing and certification is provided to demonstrate that this is fit for purpose, until then we do not intend to install cell heaters in our batteries. I know we’ll always get some flak for posting this but honestly this is not about pushing our products. The XTREME range is a niche product and we currently only sell in bulk

[@robfodder5575]

@@roamerbatteries "makes this incredibly difficult to get right" nah, any design engineer worth their salt would arrive at a working solution. All a larger mass does is slow the process down, a process that occurs anyway in the environments these batteries are in. Lets ignore the Fogstar, or other, battery lack of charge inhibit.... Can you be specific about what you feel is a safety issue with heating prismatic cells in general? Is it because you believe the heat output is a constant?

[@roamerbatteries]

@@robfodder5575 Yes you're right, a vertically integrated company with design engineers would almost certainly work out a solution. And I would expect any solution to be safety tested and certified too, I believe that is the bare minimum standard required. It is interesting that other than Battleborn (who use cylindrical cells) there are no premium self-heating batteries on the market. Of all the companies that have the resources to design, manufacture and test a self-heating battery - none of them actually do. Regarding safety, I have concerns about the uneven heat distribution which is more difficult when using large prismatic cells due to their size (5kg each) and shape (no gaps between as you would see with cylindrical cells). You can see from the thermal image that the heat distribution is not even across the cells. The increased voltage on cells 2 and 3 is also clear evidence of different temperatures. The way this is currently implemented means that the prismatic cells (or parts of the cell) are being charged faster than they should be for their current temperature and this will therefore lead to lithium plating on the cells. It is widely accepted that lithium plating lowers the temperature required for thermal runaway of lithium cells and that in severe cases, the dendrites can breach the separator layer leading to short circuits. I'm not suggesting that self-heating batteries are suddenly going to combust but I believe that the way this is currently implemented in budget LiFePO4 batteries creates an elevated safety risk and I would like to see this tested and certified before it is sold as a mass market product.

[@robfodder5575]

@@roamerbatteries "none of them actually do" - That doesn't mean its not viable, how many companies have gone under due to lack of innovation. I see this all the time and infact its one of the reasons I can put food on the table (companies paying for tech to be developed to get ahead of the competition WRT new features). You may find they dont do it simply because they are mostly in markets where it doesn't matter or there simply isn't the demand for it. WRT safety concerns, yeah fair point if those manufactures, at the very least, dont update the limits in the BMS embedded software.

[@roamerbatteries]

@@robfodder5575 Yes thats true. I'm not saying it can't be done - it certainly can. Whether they will spend the extra money on developing, testing and certifying it is another question. The Roamer BMS has a heating circuit and we could develop some custom firmware to make this work better. I personally see self-heating as a hammer in search of a nail however, I spend every winter in the Alps and have never had an issue with cold temperature charging. It's installed correctly and I maintain the temperature above 10 degrees constantly while we're away (the diesel heater runs pretty much continuously on a thermostat). And for those people that are installing under the bonnet or an outside locker then self-heating isn't the right solution anyway as it'll be constantly cycling the heating circuit - this is the use case we had in mind when we created the XTREME (although this is now mostly used for industrial applications)

[@valentijndegroot1379]

I agree the implementation of the heating is done completely wrong, it could work well with some simple fixes. If you are smart, you relocate the temperature sensor, add additional heating film in the middle and test again. Does it work? Implement it in your own product. "Our product is better because the competition did something stupid" is the wrong mindset.

[@roamerbatteries]

Yes I agree it can be improved and I sincerely hope that this video leads to affected brands making changes in their products. I think problems will persist however, using big prismatic cells wrapped in heat film is always going to be difficult. The examples I’ve seen where it seems to works well are Battleborn and Tesla who both use cylindrical cells and have a much more advanced heating arrangement. And actually I don’t think heating is needed at all for this use case, if you install and maintain correctly then your batteries should never get below zero or if they do then you can warm them slowly over a period of a few hours - no need to try and hack the process. I did have some concerns about publishing content that appears to put competitor products down, that’s not our usual style or mindset. However, this is a possible safety concern so I feel it needed to be called out. Hopefully other independent reviewers carry out their own tests going forwards

[@robfodder5575]

@@roamerbatteries I understand the BMS in the video seems to be setup wrong but I'm just not buying this is an issue for prismatic cells in the way you seem to be suggesting (yet also saying some people have done it right? edit: ignore, I see how tesla do it...). What's the thermal conductivity of a prismatic cell? All that's need is an algorithm, a couple of sensors and the solution would be fine. You seem to be suggesting that heat won't propagate, that it won't work, but it will, it just needs to be managed and managed slowly. I've got too much on to spend more than 5 mins on this but just done some heat propagation calcs using thermal conductivity 0.2W/m (assuming the cell is an EVE cell) and arrived at 2 hours from 0 degrees and closer to 3 hours from -10 degrees i.e. the battery has been allowed to be completely frozen and needs to be warmed up safely to be charged.

[@roamerbatteries]

@@robfodder5575Thanks for this. Very helpful suggestions and I agree, there are probably some pretty simple ways to improve this. Maybe start by relocating temp sensors, making sure heating is distributed evenly and also as you suggest, slowing it down (or introducing a delay) so that all cells warm up completely before charging is enabled. I hope improvements are made here, it’s in the interest of all manufacturers for safety to improve. I would personally want to see more testing and certification of cell heating before this is seen as being fit for purpose however

[@myhificloud]

Can the BMS be programmed to trickle charge when near zero, slowly ramping charge rates until various sensors determine it is safe for fast(er) charging? Essentially heating the batteries internally through a set of pre-determined charge parameters?

[@roamerbatteries]

Yes it’s possible but tricky as most BMS have no way of limiting charge current, control is either on or off. If you have Victron chargers and a Cerbo GX you can program this externally with Node Red.

[@NeverTakeNoCut-offs]

This is a GREAT video. One thing i've learned is that insulation is your friend.

[@seymourpro6097]

Is there a thermodynamic heating effect when LiFePO4 batteries are either charged or discharged?

[@roamerbatteries]

Yes, it’s a chemical reaction so it’s going to warm up when discharging anyway. And as this is the same across all cells, the heating effect will be evenly distributed - unlike the self heated batteries which only heat certain areas

[@rkan2]

It would be cool for batteries to have a ECU to calculate the probable energy required to heat up the cells for safe use (mostly the charging). It could calculate the simulated temperature of the battery and determine what energy content is required to heat it up. Shouldn't be that difficult to figure out if you have internal and external sensors and know the coefficients and masses.

[@roamerbatteries]

That would be cool. If you have Victron chargers and a Cerbo then it is also possible to program charge rates by temperature (using DVCC feature), that’s something we’re playing with now. I’m not sure if needs to be that complicated however, as long as low temp protection is in place and you have enough power to run an external space heater (to warm it slowly) then there’s no need for heaters at all. Cell heaters are ‘solving’ a problem that doesn’t exist for the vast majority of people

[@rkan2]

@@roamerbatteries The closer to an automotive application and the north pole you are, the more important it is. :P

[@Foxfried]

Sunfunkits solved this with their batteries as you can set them to maintain temperature even they are not being charged.

[@Brayco864]

You are correct sir.

[@matthewwakeham2206]

Many people won't experience -18C regularly so how do heated batteries perform at -5C? Also how do heated batteries perform when actually in use ie being charged and discharged and does that help keep them warm?

[@roamerbatteries]

The effect is the same, its just quicker... the outside of the cell pack is warmed up by the heaters but the inside stays frozen. If you're using the battery then the chemical reaction will warm it up, yes. But if you're using it in a camper then its not going to get cold anyway so the heaters wouldnt be needed.

[@rkan2]

January-March above 60N, you will more than likely experience -18 or lower :P

[@dougjohns5115]

@@rkan2 we do

[@olafschermann1592]

Yeah - you could only build the battery yourself with more intelligent architecture. Either combining more temp sensors especially in the center - or heatmatts between each cell.

[@roamerbatteries]

Yep. A good start would be to put a temp sensor between the cells. Also a good idea to program the BMS so that it only starts charging when all sensors are above zero

[@OskarHartmannsson]

the problem with more heating matts is where to locate the sensors and with prism cells they are just have to much mass to relay on temp sensors alone and adding delay means a long time without charging

[@olafschermann1592]

@@OskarHartmannsson you are 100% right. An 15kw Solar battery has a weight of 110kg / 220lbs - takes some hours to heat in case you don‘t use 1.21 GW to heat. Insulating would help a little and using solar you have some hours a day..

[@roamerbatteries]

@@OskarHartmannsson Yes true but really, if its installed correctly then dropping below zero should be a very rare occurrence. So waiting a few hours is not a big deal - you can just let it warm up naturally, there's no need to try and speed it up using cell heaters and as we've shown then this can cause more problems than it solves. If its likely to drop below zero regularly then this type of self-heating is pointless anyway, you would need something that can maintain temperatures constantly.

[@Brayco864]

Ahhhh BUT there are heat pads available that go between the cells which heat up the core of the batteries. However there are batteries that don’t allow the battery temp to get below set points ( say +4c ) so if looked after properly the battery should never sit at any - minus figure for any extended periods. If these batteries are critical they should be “ looked after “ like your laptop or mobile phone. I live in the uk so it’s unlikely I’ll ever get my Overlanding battery ever below 0c and if I did I’d be treating it with kid gloves ( excuse the pun ) before use

[@roamerbatteries]

Yes there are heat pads that go between the cells, apparently the brand we tested uses these on some of their larger batteries. This still isn't ideal however especially when heat is only applied on one side of the cell, temp sensors are located externally and a single temp sensor is used to control low temperature charging - this is still going to lead to uneven heat distribution. The batteries that don't allow the temperature that drop below the set point are a different design and have their own issues (they can discharge the pack unless you have a constant source of charging). If you're overlanding in the UK then just maintain it properly by ensuring it is stored at above 50% state of charge, and if its ever too cold then you'll always have enough energy to discharge and warm it up

[@Brayco864]

@ The Sun Fun Kit I have ( and just recently built ) the BMS monitors temperature and switches on and off the heating pads if it sees temp below a set temp that you can adjust in the software on your phone or tablet. It would automatically stop the battery getting too cold which makes putting the battery in a deep freeze experiment a bit meaningless in most cases. I guess it depends on your application BUT a battery for arctic conditions to be sat at -25 for months isn’t really what most consumers need. If I’m say overlanding in Norway for a two week stretch then stopping the battery getting too cold overnight will be beneficial as I can then recharge as I drive again in the morning. I agree that the heating isn’t the ideal design for say EVE 314 MB31 cells however I’ll do some basic testing to see what the current draw is and how well the pads work in real world conditions.

[@dougjohns5115]

I live with lithium batteries in Canada... only thing I've been able to count on 100 % is a heated battery compartment complete with temp sensor that I can read on my control panel

[@roamerbatteries]

Yep if you can maintain temperature of the battery compartment then that works well. Most people in Europe just install in living space of the van so it’s going to be warm enough anyway.

[@msturtz]

I don't know how Battleborn or others work... But supposing they add additional heat surfaces between the cells -- and monitor the cell voltage for the condition, and backing off the charge current when it happens? Also, obviously this only applies when starting out cold... But in a van, or let's say a mountaintop radio installation, the battery can simply keep itself warm all the time, rather than being allowed to get cold. Right? Eventually the power will run out, but if it's planned right, the sun will come up the next day and provide more power.

[@roamerbatteries]

Battleborn use small cylindrical cells so this is not the same issue. I expect the internal engineering is significantly better too. Yes you’re correct, self heating batteries are only designed to recover a frozen battery. There are other options out there that include a maintenance heaters to keep overall pack temperature above a set point. There are also low temperature cells such as the Roamer XTREME battery that can be charged at -30 without any heating at all

[@DTSquared]

There are certainly engineering mistakes in how the example heating system was engineered. The most obvious of which was that it seemed only use one of the three available sensors to determine that it was warm enough to safely charge. But I'm wondering if the test is representative of the likely use case. The minus 17 degree freeze was extreme compared to many installations. But more importantly I would expect that most batteries are installed and charged and therefore calable of keeping themselves warm. If the BMS was programmed to do so. So the better test would be to see if a battery could keep itself above feeezing to begin with. Or if the test battery was charged before it went into the freezer then the real issue is that the BMS was not programmed to avoid itself getting frozen. Ideally the BMS should heat the battery to keep it above zero, then lock it if for some reason it falls too low and notify user to defrost it safely.

[@roamerbatteries]

Yes there are some very simple ways to improve the design, for example moving one sensor between the cells and waiting till all sensors show above 5 degrees. The test started at minus 12 which is going to be a rare situation in the UK but not unheard of. And these batteries are used all the world. I think there’s scientific value in testing with various cell sizes and at various starting temperatures but the principle is going to be the same. The problem is caused by the heaters forcing the pack to warm up too quickly, leading to uneven heat distribution. There are leisure batteries out there which maintain temperature above a set point and this is standard on grid scale batteries. These solve the problems highlighted in this video but they are much less common because permanent heating will drain the battery (the tested battery only uses heat energy when charging so it will stay frozen until plugged in)

[@OskarHartmannsson]

it really coms down to time if you need 2h to start charging its not much use is it

[@-PORK-CHOP-]

Good video except for a few things, you both talk very quick with poor audio making it very hard to understand what you are saying. Alex was particually hard to understand

[@roamerbatteries]

Noted, thanks for feedback. There are subtitles available on the video if needed 👍

[@myhificloud]

@8:21 Hello Alex! Looking forward to you getting blown around the frozen tundra again this year. Safe travels, yaarrrr!

[@wilker1978]

Another great video ! Thanks Steve.

[@neilmick6778]

You've reviewed one battery of your choosing..

[@fitmydashcam]

One that's very popular, and indicative of what is available on the market...?

[@robfodder5575]

@@fitmydashcam Which one is it? Fogstar?

[@roamerbatteries]

We can’t confirm the brand, sorry. It wouldn’t be fair to them without giving them a chance to respond. Also, I have concerns over ALL self-heating batteries that use heat film wrapped around prismatic cells that is activated while charging, it is not specific to this one battery. As mentioned in the video, there are good examples of self heating out there. And there are simple ways this can be improved (relocate temp sensor etc) but wrapping big prismatic this way is always going to lead to uneven heat distribution

[@riohargrave]

Great video guys can’t wait for our system.

[@multicast01]

A solution can be found for low temperatures. However, I am curious about how it will be at high temperatures.

[@roamerbatteries]

The temperature threshold required for thermal runaway in a LiFePO4 cell is much higher than you’d ever see in reality. The BMS will shut down the battery way before this point

[@LeisureBit]

Really interesting - thanks for sharing - all the best, David 👍

[@roamerbatteries]

Thanks David!

[@leemobai762]

That music is annoying

[@john147GTA]

I took your advice on this very case when buying my battery, love the smart460 battery and would recommend roamer to anyone. so far I've never needed to plug into shore power

[@BlueFix]

👏🙌🫡

[@RandomDanStuffses]

This is such a clear demonstration, good work sharing this in such a clear an concise way! Glad you got the frozen chicken analogy in at the end too 😂

[@roamerbatteries]

It was only a small battery or I’d have gone full Turkey! 😂

[@activat20]

Proof of what you guys were explaining at Vanfest

[@roamerbatteries]

Yep, we finally got around to doing a proper test 😊

[@gary-xj4yc]

That’s why I’ve purchased a Roamer battery, definitely a product you can trust 👌

[@roamerbatteries]

🙌

[@JW-urth]

Heating this type of LifePo4 batteries is of course not intended for charging. More before discharging/using it in colder weather, just as you would preheat an engine. You only get more energy out of it because the battery is already a bit warmer instead of using a cold battery in a cold environment. You can forget the whole loading story and never do anything below zero in practice

[@kevinmills5293]

Your test method isn’t valid.

[@roamerbatteries]

Not sure how else we could have done it but happy to hear your ideas 🙂

[@kevinmills5293]

@@roamerbatteries perform the test while the battery is in the freezer. It would be more representative of real life conditions.

[@picobyte]

The car is hell for batteries. That's why lead acid runs the world to most electric cars for their backup.