Are used electric cars cheap because of battery degradation? How to test your EV state of health SOH

[@janhenkins]

This MG5 of yours was a good buy. My example (almost identical to yours) was hammered as a taxi, and did 43k miles in the first 3 years. In spite of that, my battery SOH is still 98.3%. This helps proving that EV battery fears are totally unfounded in general. It is far easier to find out the true state of an EV, while buying a 2nd-hand ICE car can be a big gamble.

[@UpsideDownFork]

I'd be happy with 98.3%. Avoid DC charging whenever possible and the battery should outlast the rest of the car by the looks of things!

[@johangustavsen6962]

To check the battery i would say you would have to charge to 100% and let the car balance the cells. Then you need to drive the car to 3-5% and check if you have any weak cells. Usually the weak cells dont appear before the voltage drops significantly.

[@UpsideDownFork]

Yes, that is part of the SOH calculation made by the BMS.

[@aligilani8880]

Absolute legend. So helpful 👊🏼

[@UpsideDownFork]

Glad it helped!

[@Lewis_Standing]

I couldn't get mum's MG5 to work with my OBD and car scanner But my Kona at 46000 miles is still 100%. If I try and calculate the battery capacity from draining it from 100% to as close to 0 as I can I get 63.2kwhs from 64. So pretty close to nominal original capacity. What we don't know is , what SOH is in this context

[@UpsideDownFork]

Did you definitely change the profile on car scanner?

[@THORAC1972]

The only true way to test degradation is driving the car after full charge and going down to at least 20%

[@UpsideDownFork]

@THORAC1972 why would that be any more accurate?

[@michaelgreen5515]

Very helpful so cheers

[@UpsideDownFork]

Glad it helped

[@SolAce-nw2hf]

Unfortunately the SoH estimate of an EV BMS is often not a very accurate way to determine the degradation of individual cells. It is just a rough estimate based on the total kWh's measured in the charge/discharge cycle, so if one of a thousand NMC cells has degraded 10%, it will just show 99.99% SoH But the estimate is not totally useless because with a good battery pack and decent thermal management, all cells will degrade at about the same rate for a long time. With NMC your EV usually has thousands of individual cells (look like bigger AA batteries). The LFP EV cells are usually much bigger and box shaped and are often monitored and balanced individualy by a more advanced BMS. When one cell starts to age prematurely, it can be very easy to spot because of the voltage curve during top balancing. Also a smaller number of bigger cells make it easier to swap out for a good one than trying to tear apart hundreds of smaller NMC cells tightly packed into the EV.

[@UpsideDownFork]

Thanks for the comment. When I first got the car I charged it to full with the slow granny charger to allow it to do the top balancing. This should have given the BMS the best chance at being accurate, but I do accept that there is room for a margin of error. My MG5 battery, like many others, has a buffer of ~3kWh built in so that the battery always has a little safety zone.

[@simonhenshall4294]

Would capacity in amp hours be more useful? Is this something that a lay person can access? Any data relating to EV battery capacities over time and with use that you can signpost to? I’m particularly interested in the Mercedes EQC battery.

[@UpsideDownFork]

@@simonhenshall4294 Amp hour numbers are accessible on some vehicles, dependant on the BMS. I think on Mercedes you will only access that data through Xentry Star. Whilst I have that kit, your average person or even most garages will not. The EQC is fairing well, it's been around long enough to see several cars with high mileage. Generally, the BMS on those seems to be managing the life well. This is a good sample of Tesla data. kzbin.info/www/bejne/noXXY5-fl7RgrK8 I expect the EQC to be broadly similar as the NMC chemistry is equivalent and the BMS on both operates in the same fashion.

[@RBcymru]

Totally agree. These scanners measure voltage and guess. I commented above.

[@UpsideDownFork]

@@RBcymru these scanners do not measure voltage and guess. The BMS measures voltage, load, discharge and charge rates and reports it. The scan tools are just a way to access that BMS data.

[@kevinmatthews2620]

spot on, i would like to know where you got your mg5 from that would suit me 100% especially @ that price :)

[@UpsideDownFork]

Set up automatic searches on eBay and auto trader and sit back and wait for the adverts to come to your email inbox. Good deals don't last long.

[@wisamrab2724]

Where can i get this check piece ?? Is it on aliexpress ?? Please help 😊

[@UpsideDownFork]

Did you read the video description?

[@richardpratt5251]

Thanks for this video. Unfortunately, when I select CITIGOe-iV the ‘All Sensors’ screen is very different to yours. In particular, there is no entry for ‘State of health' below ‘State of charge’. There is an earlier entry called ‘Battery power loss’. The value given is ‘0’, and I wondered if this might be equivalent to a health of 100%?

[@UpsideDownFork]

Each car will be slightly different and it's been a while since I checked an e-up! but I did think it had a normal SOH field. Otherwise I would advise seeking help from a specific owners group who might know about any peculiarities your car has.

[@MyfirstEV23]

It's seems the MG 5 has a better battery technology, I have seen cleverly EV MG 5 holding up really well but I can't say the same for the ZS ev models. My car is 2021 model and only 48000km my SOH is at 90.65% which is really a shame. My 2011 Nissan leaf with no thermo management system was 82 % in 2024 before I sold its for the MG zs ev due to CCS and thermo managements but the Zs ev model has really let me down.

[@UpsideDownFork]

@@MyfirstEV23 that's a shame to hear. Do you rapid charge often? I've spoken with a good few MG5 taxi drivers and most of them have not noticed any battery degradation after a couple of years.

[@MyfirstEV23]

@UpsideDownFork No, I have really taken care of the battery ,I have charged the car 98.9% at home with a slow wall box around 3kwh, and I try to balance the cells regularly. I charge 100% when I need the range. I live in Sweden and not in hot climate so there wasn't any reason for the massive degradation. I bought the car from a former lease in June 2024 ( 3yrs old) and 32000km. SOH was 92.7% already.

[@UpsideDownFork]

@@MyfirstEV23 Ah, that's a shame. I'm sorry to hear that.

[@robwoodphotos]

Check out James and Kate videos. James heads up Cleevely motors mobile ev servicing network, which kit out MG5s for their mobile mechanics … he raves about MG5s reliability … he has done 158,000 or so in his and it’s still 91% … plus still hanging together well as a car

[@UpsideDownFork]

Their videos were a big part of convincing me to buy this car 👍

[@robwoodphotos]

@@UpsideDownFork same here … I use cleevely to service my two EVs and they are terrific

[@RBcymru]

No you need a load test not a voltage test.

[@samiaksu4005]

My vehicle is standard mg4. When I bought it, SOH was 93.64% and 10500 kilometers. It is currently at 21000 kilometers and SOH is 92.80%. So I see a loss of 0.8% in 10000 kilometers. Also, I don't know how the first owner lost so much.

[@UpsideDownFork]

@@samiaksu4005 thanks for sharing. Perhaps the first owner only used rapid chargers.

[@samiaksu4005]

@UpsideDownFork Maybe. But maybe he only drove with a high percentage of charge. So he usually drove between 80% and 100%, or waited for a long time with a high percentage of charge in hot weather. There are many possibilities.

[@UpsideDownFork]

@@samiaksu4005 Yes, true!

[@andrewtruscott8515]

Do you know if that will work with a Tesla model 3, 2021? Thinking of buying but I’d like to do this before purchasing…

[@UpsideDownFork]

This is what you need to watch as Tesla is a little more difficult than other legacy manufacturers. kzbin.info/www/bejne/hJXdk2Z9iM2FisUsi=GSmM-o44tJMAVtQU

[@RBcymru]

No use the Tesla test inbuilt in the car. It charges the car to 100% then uses aircon etc to load test the battery. It i better but a battery can fail the day after, Look at the charging history or ask Tesla for it. If the car has been on rapid chargers more than twice a month don't buy the car.

[@crm114.]

Used EVs might be cheaper because of concerns over battery degradation but in all but the early Leafs which had poor thermal management, it’s really a non issue. Teslas with 100,000 miles on the clock have an average of just 12% loss of max charge. My 4 year old Model 3 has lost just 3% of its initial range. Unless you’re unlucky (in which case you’re protected by the warranty) batteries are good for at least 200,000 miles and LFP batteries even longer.

[@paulrnaylor]

Unfortunately the very people who will buy 10 year old ev are the ones who can't afford to risk big money on buying it with duff battery. Only time and experience will change people mind.

[@robinbennett5994]

Yeah, used EVs only seem cheap because they were over-priced initially and every year the new ones get better and cheaper.

[@UpsideDownFork]

Yes, agreed. We need to keep banging that drum that EV batteries are not as scary as you think.

[@robinbennett5994]

@@paulrnaylor a 10 year old Nissan Leaf is about £3k on Autotrader these days. At that price it'll pay for itself in 2 years. It doesn't have to be a long-term investment to be good value.

[@_Dougaldog]

It looks like your purchase will serve you well, is that LFP or NMC battery ? Some of the data looks a bit odd, speed data as one example, do you have confidence in what you're seeing ? Would you have any recommendations for OBD dongle, in terms of functionality which would be best ? Apologies for all the questions (Geek mind in overload) :-)

[@UpsideDownFork]

This is the only dongle I personally recommend. amzn.to/3TovmC4 It will do anything, on any car. I've never had it trip up on any vehicle, with any protocol. The app is the limitation. I always leave a dongle in every car I own. I've been in a position to help many people over the years by doing so. In theory the battery is LFP, but I have seen some evidence online that some Europe market cars may have been fitted with NMC packs. I initially thought this was just the difference between my 52.5kWh and the later 61.6kWh, but now i'm not so sure. I need to jack up the car and check the labels on the pack itself to know for certain. Unfortunately these free apps do sometimes miscommunicate through certain OBD protocols and the data is not always perfect. Less mainstream cars can sometimes be more susceptible to these glitches. I may try one of the other OBD apps and see if those sensors read differently. Typically you get a number within the normal range or it is well outside the typical scope and can be disregarded. In my case I have access to professional scanners from Snap-on, Autel and other dealer level tools. I have checked my car with the snap on scanner and it does come back as 100% on that too.

[@_Dougaldog]

@@UpsideDownFork Thank you for the comprehensive reply, no stones left unturned there 👍 One OBD dongle now ordered, cheaper than I was expecting. I wonder, to identify battery type, could the charge settings screen give indication by default ? For example my charge screen (0-100% green line) on my MG4 extended range NMC battery type, has a 'battery health mode' (40-80%) and 'long journey mode' (80-100%) annotations above the slider. Which give a general indication as to the best levels of charge for battery health. Would I be correct in thinking the LFP battery models might not have these markings (as per P147 of MG5 manual) ?

[@stewreviews9345]

I like these but am unsure on how "health" is measured from car to car. Is it always the same? I've seen people checking old leafs and getting a variety of scores that reflect the predicted mileage but generally we now just see 100% or thereabouts. Clearly battery tech has improved hugely, but equally battery degredation hasnt just gone away - do you often find vehicles with lower health scores? Paradoxically the reassuring video for me would be the one saying, oh dear, this doesn't happen often - SOH is 78%, best avoid this one...

[@UpsideDownFork]

100% at 30k is rare in my experience. I would expect 97-98% typically. The leafs do not have any active thermal management, that's why they are the cars with much higher degradation rates. Along with this, they used the NMC chemistry which is prone to higher degradation rates, as now shown in vehicles like the Tesla Model 3. LFP batteries are doing much better than their NMC counterparts. Add to this the fact that people are understanding that AC charging rarely does or can do much damage, as well as limiting SOC to 80%. DC rapid charging and regularly charging to 100% are the main drivers of battery degradation. In terms of how health is measured, well if you want to get into the weeds, you will actually find that a lot of manufacturers build in buffers. On this MG5 for example, the pack has a Nominal Capacity of 52.5kWh but the accessible capacity is 48.8kWh, a reserve of 7%. Some say that the first 7% of degradation will never be reported because it is not accessible by the BMS. The long range version of this car has a pack of 61.1kWh but 57.4kWh usable. Also a buffer of 3.7kWh. Once again, earlier cars didn't build in this reserve buffer. Massive overgeneralisation here but the Kia/Hyundai vehicles are a real sweet spot. They manage their batteries incredibly well and it's not unusual to see cars with over 100k miles and still 97% SOH.

[@Glyn001]

👍👍👍

[@UpsideDownFork]

Thanks 👍

[@SuzhouChen]

Personally I don't trust this data very much, I think the proper way is to look at the voltage differences between cells, however, that is not something we can do without taking something apart and get deeper into the vehicle

[@UpsideDownFork]

Watch again. The BMS does report the cell by cell data. In this instance every one of them was 3.21V

[@SuzhouChen]

@@UpsideDownFork Sorry I should watch all before replying :D

[@rocket3man]

I don’t have that issue on my 41/2 year Mini and 4 year old Tesla. If you drive EV’s you don’t need this nonsense.

[@UpsideDownFork]

@@rocket3man did you buy new or used? If used why didn't you test the battery?

[@rocket3man]

@@UpsideDownFork new for both, but with the experience of batteries that they do not stop working after a few years what is the need to test. Get a full charge see what the lie-ometer tells you, look at the Wh/h average is and multiply by you battery capacity that will give you a clue (a big one) of the battery condition.

[@UpsideDownFork]

@@rocket3man Unfortunately not everyone has the same confidence in EV batteries so we need to collectively dispel the myths around degradation. I can check the SOH on any EV apart from a Tesla in under 60 seconds. This is a quick, cheap and easy way to give potential buyers confidence in second hand EV's.

[@RBcymru]

Total rubbish because every EV battery will "settle down" in the first 12 months and not be 100%. Tesla do a test where the car is fully charged and then load tested by the car to check battery health until the battery drops to a low state of charge. Your 3 year old MG has Lithium Ion batteries pretty much the same as your phone. In your 3 year old phone check battery health and like mine which is at 89% your car will have lost 5 to 10%. The test a lot of these so called scan tools is on Voltage which can be OK but when a load like a motor is drawing current the voltage drops. Its the same in AA batteries the voltage might be about 1.5V but under load might drop. Your EV will say 100% charged but 100% of what ? 100% of 50 kw or 100% of 45 kw ? Your range is a guess calculated on recent driving so you need to check the long term efficiency. If you charge at 3kw your degradation will be less than if you charge at 7kw and rapid charging will degrade your battery even more. Every battery degrades with age so your reading is wrong.

[@UpsideDownFork]

Nice. A phone battery is not necessarily directly comparable to a car battery due to thermal management and more advanced cell level BMS data. These scan tools don't produce the reports the BMS does. To get an accurate reading, you will want to perform a cell or top balancing charge at least once within the past 30 days, otherwise the reading can be 1-2% out. Most Tesla batteries are NMC chemistry rather than LFP so they do experience more degradation. The real elephant in the room here is the in built buffer which is not available to owners. That's the part of the battery that is slowly being lost and why many cars still report 100% after many miles. The usable part of the battery is the same at new or at 30k miles and we're now seeing this to be true pretty much across the board as OEMs are being more conservative with lots of things like thermal management and buffers etc. Finally, the difference between 3kW and 7kW AC charging. Tests have shown that most cars prefer the 7kW charging because the AC-DC charger that's built into all cars is operating more efficiently at that rate. I agree that DC rapid charging does no favours to the battery though.

[@RBcymru]

@@UpsideDownFork LFP batteries don't like cold weather and still take longer to charge from 80 to 100%. The 800 V battery packs will suffer more because they are rapid charged more often. The BMS isn't that accurate at about 2%. The 7kw comment doesn't really stack up because the car is drawing twice the current. The longer a battery takes to charge is better for its long term health. The car charger should be able to handle 3 or 7 kw of charge but if they like higher rates is that the design of the charger and battery ? Are manufacturers designing cars to only last 8 years so we buy more ? It wouldn't be the first time car makers have designed crs to last a set time.

[@wisamrab2724]

Can I get your email to contact you ?

[@UpsideDownFork]

Did you read the video description?

[@TheAffiliater]

Surely No one is still falling for buying the electric Car Fad. 10 times worse for the environment, so much heavier! Risk of burning down your house and fugly too! Not to mention how to dispose of old Batteries?

[@UpsideDownFork]

10 times worse for environment? Incorrect. Much heavier? Incorrect. Risk of burning down your house? Incorrect. Ugly? Subjective but most EVs are also available as petrol versions so not sure how this is relevant.