Hyundai IONIQ Forum banner

201 - 220 of 228 Posts

·
Registered
Joined
·
124 Posts
Reason I said 3%, is that I'm assuming the battery pack of the Soul 30kWh is the same as the Ioniq28kWh, which is close to 31kWh.
The BMS is a software algorithm used in a computer that measures the voltages and temperatures of the battery pack.
Having the same BMS is completely independent of the battery pack.
The Kia Soul 30kWh and the Hyundai Ioniq 28kWh have completely different battery packs.
Made by different companies and with very different chemistries. Buffer sizes and total capacity are as yet unknown for both.

pmiddeld - try using the soul spy app to show the battery cell voltage heat map. I suspect one of your cells may be going bad.
 

·
Registered
Joined
·
28 Posts
Today I drove the car from 100% to 2.5% (screenshots attached). I drove the same route I did about 3 years ago when I have had the car for 3 months, but this time the temperature was at least 5-10 degrees lower, and I was using winter tires. I estimate the car drove about 15km less now compared to 3 years ago, but I will try to repeat later when temperatures are higher and summer tires are back on.

Starting voltage was 396.9v, which is average 4.134v.
I drove halfway, took a screenshot, parked the car for 2 hours while taking a walk and dinner and drove home again.
On the way back I stopped at 2.5% to charge, at which point I had driven 222km with usage of 10.9kWh/100km. It is closer to 3% than 2.5% (car just went from 3 to 2.5% as I drove in to charge). Est. capacity = 2.22*10.9 / 0.975 ~= 24.82kWh.
However, if I look at Discharged/Charged totals, Discharged total is 12287.9kWh - 12257.1kWh = 30.8kWh, and Charged (regen) is 12662.8kWh - 12658.5kWh = 4.3kWh. So used 30.8kWh but gained 4.3kWh = 30.8kWh - 4.3kWh = 26.5kWh taken out of battery from start. (The numbers here are +/-1% in charging and discharging losses, but I'm assuming they are evened out). The BMS is 95% at 100% charged, and at bottom it was 4% left on BMS at 2.5% SoC. So that means I used 91% of the total BMS covers. 26.5kWh/0.91 = 29.12kWh. Estimated new capacity of the battery is around 30.96kWh 29.12kWh/30.96kWh = 94.1%.
I choose to interpret this as my car having lost ~5.9% of it's original capacity?

What do you think? It would be nice to see other people try with their cars, and take note of the BMS% instead of the SoC%.


33951
33952
33953
 

·
Registered
Joined
·
28 Posts
I need you guys to help me shoot down a new theory I have. The Ioniq measurement of kWh/100km is about 5-6% off for some reason. The reason I say that, is that yesterday I pulled 30.8kWh-4.3kWh (regen) from the battery down to 2.5% left. That's 26.5kWh. But if I calculcate 222km * 10.9kWh/100km, I get 24.198kWh/0.975 for missing 2.5% = 24.82kWh. 26.5kWh/24.82kWh = 1.0678. About 6.78% difference between the cars reported leaving/entering battery, vs the kWh/100km measurement.

So my new theory is: The Ioniq doesn't measure energy leaving the battery for the kWh/100km value. It seems to measure some other value. Either the energy applied to the wheels, or perhaps, the energy leaving the DC-AC converter to the engine. Since most people trying to measure how much they can pull from the battery seems to end up just south of 26.5kWh, 26.5kWh * 1.06 = 28.09kWh, it almost seems plausible.

BUT, that means, SoC 0-100% is actually showing 28kWh available. There are no extra buffers except the visible top 5% and bottom ~2.5-3%. 28kWh / 1.08 = 30.24kWh (which is close to the estimated full capacity of 30.96kWh).

I noticed yesterday, that my car went into super-turtle mode already at 5% even though battery was 22C. A few years back, I am pretty sure it activated at 3%, which could indicate that my battery is allowing me to go down to a lower voltage than before. The bottom buffer also seemed to have shrunk to 1.5% compared to 3% 1.5 years ago. Is this the buffer that the car is eating to keep up 28kWh available?
 

·
Registered
Joined
·
237 Posts
The consumption is undoubtedly easier to measure from the DC/AC inverter and at that point there are heat losses in the system, it could explain the missing kWh.
 

·
Registered
Hyundai Ioniq premium -19, 28kWh ~46k km
Joined
·
53 Posts
Haha, funny thing! I did the exact same test two times the last week. One on three intensive short trips and one on a longer calmly driven trip. Broke my phone a few days ago so no more testing until my replacement has arrived.

The first run was down to 1.5% driving [email protected]/km = 24.746kWh/.985 = 25.122kWh. Took 27.3kWh to fullt charge again.
Second run was down to 2% driving [email protected]/km = 24.478kWh/.98 =24.978kWh. Took 27.01kWh to fully charge.

My car has done 46000km and was built in september 2018, in traffic may -19. So the battery pack should be about ~3 years old by calendar date if it was manufactured a few months before tha car. (The cells might have been batch manufactured in 2016-2017 på LG and stored waiting for assembly, we don't know)

Now the battery was under 10˚c most of the km driven in both tests, my theory is that I will get more out of it in the summer. But Ill test that when its warm enough. Consumption and distance as given by the trip and consumption meter on the dashboard so they should be synced.

here are some pics from the second run:
33956

Before the trip, full here is 94.5% bms due to cold battery(at -10 its around 90% bms = 100%soc or something, ill check my torque screenshot archive)

33957

After the trip, hit super turtle at 3% as usual.

33958

After charging back up, notice how it now charged higher than before, 95% bms instead of 94.5% before due to (presumably) a warm battery. I used 91% of the bms capacity for the trip which gives a capacity 24.478kWh/91% = 26.898kWh including the buffers.

now, the soc on these cars is not linear, the bottom 10% contain less energy than the top 10%. I think i have some german graphs on my computer I’ll try to dig up that support this.

Edit: Here is the very interesting graph!
33960

By Addur on the goingelectric.de forum, where I'm an avid lurker with google translate, but it does not translate pictures. Can't remember the thread I found it in though.

Now, what it says, I don't have a clue, I've not yet been bothered trying to translate it. But the curve(big blue is average) does speak pretty much for itself. At low soc we get about 200Wh/1% and at high soc almost the double at 380Wh/1%.

Also, that torque picture from with a battery at -10°c when the car was fully charged to 100% SOC the BMS shows 92.5%, indicating that the BMS is not allowing a full charge on a cold battery.
33961
 

·
Registered
Joined
·
28 Posts
The first run was down to 1.5% driving [email protected]/km = 24.746kWh/.985 = 25.122kWh. Took 27.3kWh to fullt charge again.
Second run was down to 2% driving [email protected]/km = 24.478kWh/.98 =24.978kWh. Took 27.01kWh to fully charge.
As I wrote in my post, I'm not sure the kWh/100km measurement displayed in the car is accurate, as I suspect it may be on the AC-side (engine) on the AC-DC converter, and not on the DC (battery). This means usage is showed as lower (factor 1.055?) and regen is showed as higher (factor 1.11?). This needs a lot more investigation to confirm.

My car has done 46000km and was built in september 2018, in traffic may -19. So the battery pack should be about ~3 years old by calendar date if it was manufactured a few months before tha car. (The cells might have been batch manufactured in 2016-2017 på LG and stored waiting for assembly, we don't know)
I believe this is exactly the same as my car, so the additional decay in available energy would probably be down to mine driving an additonal 20000km.

here are some pics from the second run:
The more interesting thing to calculate here is the Charged/Discharged (sadly you do not have the decimals, only the whole kWh, but still):
Start, 94.5% BMS, 7130kWh charged, 6901kWh discharged
Stop, 3.5% BMS, 7132kWh charged, 6929kWh discharged.

So that means you used 91% BMS of the battery (at 100% - 2%), discharging 28kWh, and charging 2kWh, so a net total of 26kWh used, putting max capacity at 26kWh/0.91 = 28.57kWh. But since you are lacking decimals on the charged/discharged, it could also be that we're +/- 2kWh, putting you at 26.4 to 30.8kWh range. We need decimals to more accurately measure it.

After the trip, hit super turtle at 3% as usual.
Interesting, I get super turtle at 5% now, and at 2.5% SoC (4% BMS) I had much lower voltage than you. Is that just the cold, or is it degradation?

Now, what it says, I don't have a clue, I've not yet been bothered trying to translate it. But the curve(big blue is average) does speak pretty much for itself. At low soc we get about 200Wh/1% and at high soc almost the double at 380Wh/1%.
This is not really unexpected, and I'm not sure if it is a quirk of the batteries or a quirk of the SoC/BMS. But the good part is how reasonably flat the curve is after 80%.

Also, that torque picture from with a battery at -10°c when the car was fully charged to 100% SOC the BMS shows 92.5%, indicating that the BMS is not allowing a full charge on a cold battery.
I experienced the same with my car. When temperature is low, the BMS doesn't go as high. But I also experienced that the car would not charge to 100% when it was cold, my theory being that the input power (charging on medium from wall) isn't enough to fill the battery AND keep the heating of the cells going at the same time. This sounds strange, but it may be a bug in the software of the BMS, thinking it doesn't have more power available than what it will allow into the cells (the charging goes really slow last 1%), thus aborting charging.

I do believe however, you need to redo the tests with a 15C battery or higher. But it is quite a loss in capacity in the winter if the BMS is protecting the battery. Important to know when using ABRP wintertime.
 

·
Registered
Hyundai Ioniq premium -19, 28kWh ~46k km
Joined
·
53 Posts
@TNSe
I've been suspecting that the car is reporting a different figure from what is drained from the battery since I got the car but never really understood it and started counting consumption from my wallcharger instead. In my early measurements i was comparing to a full 28kWh capacity and assuming soc energy content was linear.
I will keep logging a trip every now and then and eventually I should have a decent database to draw some conclusions from. What I’m really missing is the early/new measurements but I guess teslabjörns will have to do of being able to extract about 26kWh from a new car(the early demo one might have had a different bms on it)

I don't know what happened to my decimals, il have to look into it when i set it up again on the new phone, I to was disappointed when I noticed.

I would love to have a reading from when the car was new, I did a 240+km trip and got home with 2% the first week I had it... but no photos or logs of the exact distance and consumption to be found yet.

charging in extreme cold is all sorts of nasty. Battery needs heating and so does the house so the wallbox limits charge power to allow the house heater to run without busting the main fuse, to add some icing on the cake consumption also increases. I don't think it has much to do with degradation though.

Edit jackpot!: Sometimes you just gotta look a little deeper into backup folders. Found this picture dated 8 june 2019, car was practically new(8 days, 939km) and I think it goes with this instagram post of mine a few weeks later. I got stuck in an unexpected heavy summer thunderstorm coming home but the temperatures were warm, 20-25°c and dry for the first 165km(just the last 55km home were very wet), so battery should have been at summer temperatures.

Now this calculates to 26.691kWh used for a calculated capacity based on soc of 27.236kWh. Assuming SOC was BMS-SoC 3-95% it should be (26.691kWh/90%) 29.656kWh (BMS 95-5%).

Compared to this I should have about 7.5% capacity loss(24,7/26.7). How much is due to degradation and what is due to temperatures we will have to wait another two-three months or so to figure out. :)
33963


Edit Jackpot #2:
I just remembered I logged the trip too, to make some fancy graph that I never got around to but when I checked the logs they do have the decimal saved (Hurray!) so its only a display thing I need to correct.
CEC: 7129.9 -> 7131.5 = 1.6 kWh regen (mostly coming down a hill from 350m to 100m)
CED: 6901.1 -> 6929.1 = 28 kWh discharge
26.4 kWh / .91 (BMS) = 29.010 kWh

The original question ofc still remains, what was the original capacity and how much is degradation and how much is cold cells. :)
 

·
Registered
Joined
·
28 Posts
What I’m really missing is the early/new measurements but I guess teslabjörns will have to do of beeing able to extract about 26kWh from a new car(the early demo one might have had a different bms on it)
Either that, or Hyundai changed how they measure the infotainment kWh/100km value in a software update to the infotainment. I have a vague memory of the ioniq (or was it another electric car) that could show negative kWh/100km because it kept counting while charging. If Hyundai didn't have the available numbers to avoid showing negative values they might have grabbed another value which is just not correct (but within 5-8%)
 

·
Registered
Hyundai Ioniq premium -19, 28kWh ~46k km
Joined
·
53 Posts
Either that, or Hyundai changed how they measure the infotainment kWh/100km value in a software update to the infotainment. I have a vague memory of the ioniq (or was it another electric car) that could show negative kWh/100km because it kept counting while charging. If Hyundai didn't have the available numbers to avoid showing negative values they might have grabbed another value which is just not correct (but within 5-8%)
Yeah, that could be a valid explanation for how TB got out 28kWh from the demo car the first time but no one else has managed to do that, not even himself. Using the same method we use here, cars displayed consumption and distance driven, he gets numbers around where we are too.
(btw, I edited my post above while you wrote the last one too, found some extra numbers on my computer.)
 

·
Registered
Joined
·
28 Posts
(btw, I edited my post above while you wrote the last one too, found some extra numbers on my computer.)

Now this calculates to 26.691kWh used for a calculated capacity based on soc of 27.236kWh. Assuming SOC was BMS-SoC 3-95% it should be (26.691kWh/90%) 29.656kWh (BMS 95-5%).
26.7kWh is probably the maximum I've seen anyone get from the km * kWh/100km measurement. And in line with all other tests. And it is a great comparison for testing when temperatures go back to 20 degrees for a relative comparison. We can either interpret it as a 5% degradation from 28kWh (28kWh * 0.95 = 26.6kWh) or that kWh/100km is off by 5%. I am heavily leaning towards that the displayed usage is wrong.

CEC: 7129.9 -> 7131.5 = 1.6 kWh regen (mostly coming down a hill from 350m to 100m)
CED: 6901.1 -> 6929.1 = 28 kWh discharge
26.4 kWh / .91 (BMS) = 29.010 kWh
So, assuming "new pack" has 31kWh (30.96kWh) you are looking at a loss of 6.5% (93.5% remaining). I actually expect 2-3% to be lost due to the starting temperature, so when you get your pack up to 10-15C we can see if the values increase. In this measurement, your car comes out just slightly worse (0.5%) than mine which has 68000km driven. (And quick charged 80 times norway-spain and back, many times with 30C in the shade in France). Wish I had OBD2 dongle for that trip...
 

·
Registered
Joined
·
124 Posts
JejuSoul I don't think I need soul spy for that. See the pictures. It is not just one cell that is bad.
Those numbers look fine for me.
It looks identical to a recent test I did on my car, which has a 'new' battery.
Cell voltage deviation was zero at all times except for the very lowest SOC when it reached 0.04V.
see Analysis of a replacement battery.
 

·
Registered
Hyundai Ioniq premium -19, 28kWh ~46k km
Joined
·
53 Posts
26.7kWh is probably the maximum I've seen anyone get from the km * kWh/100km measurement. And in line with all other tests. And it is a great comparison for testing when temperatures go back to 20 degrees for a relative comparison. We can either interpret it as a 5% degradation from 28kWh (28kWh * 0.95 = 26.6kWh) or that kWh/100km is off by 5%. I am heavily leaning towards that the displayed usage is wrong.
Actually, degradation might be the answer here too. I know my battery was atleast 8 months or ~240 days old when I first got ownership with 13km on the odo, probably older. According to this paper(think ive linked it before, page 65) the calendar degradation at 25˚c could be upwards 5% in this timeframe, especially if the cells predate the car by a significant amount. Then again if stored correctly and were manufactured the same time the car was it could be closer to 1% degradation.

Now, I agree that something is up with either the displayed consumption or the CEC/CED numbers since they dont align. It would be nice to see the label on one of the cells, for a date and capacity/model number stamp.

So, assuming "new pack" has 31kWh (30.96kWh) you are looking at a loss of 6.5% (93.5% remaining). I actually expect 2-3% to be lost due to the starting temperature, so when you get your pack up to 10-15C we can see if the values increase. In this measurement, your car comes out just slightly worse (0.5%) than mine which has 68000km driven. (And quick charged 80 times norway-spain and back, many times with 30C in the shade in France). Wish I had OBD2 dongle for that trip...

This does sound reasonable, we will see when it gets warmer. Mine has about 450kWh of DC charging done on about 30 sessions. Its never been out of Sweden, yet. I got to Öland, 550km as the furthest yet, its almost abroad but not really. :)
 

·
Registered
Joined
·
28 Posts
Actually, degradation might be the answer here too. I know my battery was atleast 8 months or ~240 days old when I first got ownership with 13km on the odo, probably older. According to this paper(think ive linked it before, page 65) the calendar degradation at 25˚c could be upwards 5% in this timeframe, especially if the cells predate the car by a significant amount. Then again if stored correctly and were manufactured the same time the car was it could be closer to 1% degradation.
But this paper seems to look at NMC/LMO batteries? Ioniq has LiPo batteries. LG Chem has this paper on their LiPo batteries: (Not confirmed this paper is the exact battery in the Ioniq, but it matches on volt, Ah and form, L3 LG Chem - LQ 1729-A2, if someone could weigh one cell, it should be 966g +/- 12g)
33965


My car seems to follow the graph at around 10C. (About 400 cycles, 94% remains). The spanish car earlier with 1400 cycles seem to follow the 25 degree path. So what does that mean for me? At about 800 cycles (~130k km) my SoH will probably drop below 100%. The LiPo battery in the Ioniq is not very fond of the nordic climate.
 

·
Registered
Hyundai Ioniq premium -19, 28kWh ~46k km
Joined
·
53 Posts
@TNSe yeah, li-ion batteries is a whole lot of a complicated mess and lots of short/slang etc.
The polymer part just denotes that the electrolyte is a polymer. Then you have an anode and a cathode too(and some other separators and packaging etc). The battery in Ioniq is probably a NMC111(or possibly NMC523) cathode with a graphite anode and polymer electrolyte.

The LG test is also a very agressive 100% Depth of discharge at 1C(full discharge(280Wh/km or charge(28kW) in 1 hour), so degradation should be (a lot?) higher there than in reality with a real driving cycle(the chalmers paper has lower DOD on most of its tests, but higher c-rate(to simulate fast charge)). My typical driving would be a 0.4-0.5C(110-140Wh/km) discharge and 0.1C(AC 12A, 2.7kW) charge for instance.
(C-rates are measured in Ah, so converting to energy(Wh) is not exact without taking account for pack voltage, but close enough here)

Types of lithium-ion polymer batteries
When we talk about different types of LiPo batteries, we are referring to batteries that modify their positive electrode materials to achieve different energy densities, capacities, and temperature resistance; depending on the material, they can be used in different applications. These are also known as different chemistries. The four main chemistries for LiPo batteries are the following:

  • Lithium cobalt oxide battery (LiCoO2, LCO battery)
  • Lithium-ion ternary battery (NCA, NMC battery)
  • Lithium-ion manganese oxide battery battery(LiMn2O4, LMO battery)
  • Lithium iron phosphate battery (LiFePO4, LFP battery)
Sourced from here
 

·
Registered
Joined
·
28 Posts
@Xiziz You said you had logging, just how much is logged and how often is it logged? I want to log specific values while driving and set up some automated interpreting to see just how temperatures, SoC and so on interact.

Also, TeslaBjørn has an Ioniq that charges at near 69kW all the way to 77%? Mine stops at 75%. Maybe I'm too used to winter these days. I gotta check again when it's summer.

 

·
Registered
Hyundai Ioniq premium -19, 28kWh ~46k km
Joined
·
53 Posts
@TNSe I use torque pro and a cheap chineese elm 327 clone. I can log any value torque displays every half second or so, but the more different values(power draw, speed etc) you add to the logg the slower it seems to become. I think there are fast obd dongles that can do 60+ readings per second. But its really just a matter of what your logging, if logging power draw to determine average consumption you need a fast time, i think ill set mine back down to 5sek intervals in the future, had it fast when testing my aero inserts for my winter rims.

I have it set to manual log and start/stop it when i want to log something. I dont drive with the obd connected normally, only when testing or if i want a screenshot of the values ill connect it just for that.

Log excerpt in the spoiler
Device Time000_Battery DC Voltage(V)000_Battery Power(kW)000_Cumulative Energy Charged(kWh)000_Cumulative Energy Discharged(kWh)000_State of Charge BMS(%)000_State of Charge Display(%)003_VMCU Ambient temp(°C)003_VMCU Real Vehicle Speed(km/h)004_CALC Average Battery Module Temperature(°C)004_CALC Average Cell Voltage(V)
15-Apr-2021 11:03:42361.312.837130.36908.57275.5787.536.253.76
15-Apr-2021 11:03:42361.312.837130.36908.57275.5787.66.173.76
15-Apr-2021 11:03:43361.312.837130.36908.57275.5787.66.173.76
15-Apr-2021 11:03:43361.312.97130.36908.57275.5787.66.173.76
15-Apr-2021 11:03:44361.312.97130.36908.57275.5787.566.333.76
15-Apr-2021 11:03:44361.312.97130.36908.57275.5787.566.333.76
15-Apr-2021 11:03:45361.312.97130.36908.57275.5787.566.333.76
15-Apr-2021 11:03:45361.312.837130.36908.57275.5787.566.333.76

Edit: Ive seen 69kW a few times and mine usually throttles at 77% but ive noticed its very dependant on battery temperature, sometimes it throttles at 75% when the battery is colder(like the first dc session on a long trip).
 

·
Registered
Joined
·
28 Posts
I've seen 69kW a few times and mine usually throttles at 77% but ive noticed its very dependant on battery temperature, sometimes it throttles at 75% when the battery is colder(like the first dc session on a long trip).
Yeah, I have seen it several time too, but not the last year. But I wonder if it depends on the charging station as well, I've seen that these Delta 150kW chargers around here be very conservative. When they originally came, they didn't even work with my car. Operator said it was a data error originating from the car that caused the incompatibility. On Ionity I also sometimes experience charging just stops, at exactly 53%. But last time I used Ionity, the charging stopped, then a few seconds the process automatically started over and continued.
 

·
Registered
Joined
·
124 Posts
...
But this paper seems to look at NMC/LMO batteries? Ioniq has LiPo batteries. LG Chem has this paper on their LiPo batteries: (Not confirmed this paper is the exact battery in the Ioniq, but it matches on volt, Ah and form, L3 LG Chem - LQ 1729-A2, if someone could weigh one cell, it should be 966g +/- 12g)...
The first paper you refer to is the Chalmers article on battery degradation. That is a really interesting article. But it is not the Ioniq battery cell.
I think the conclusions apply to all cells however. Do not leave an EV at high SOC, particularly in very hot weather.
The problem for us is that dealers invariably leave all their new cars at 100% SOC all the time.
Hence it is far better to buy a new car in the Spring rather than the Fall.

The second paper you refer to is the spec sheet for the L3 LG Chem - LQ 1729-A2 cell. Your graphic comes from this spec sheet.
I linked that paper on this site 3 years ago. But the link is now dead.
New battery layout and motor components in Hyundai / Kia Electric Vehicles.

Instead you can currently see the spec sheet here - Chinese Site selling LG Chem – LQ 1729-A2 Cell
 

·
Registered
Joined
·
15 Posts
Hi folks!

I'm going to look at a used 2017 EV Ioniq next week and I've read every post I can in this forum as it's off a dodgy dealer and the main thing I'm worried about is the battery.

What I'm getting from all these posts is that the batteries are doing really well in these cars.
Also that if I get the (simplest way to do this) Soul EV app and check the SOH and see it's better than 95% (probably higher going by these posts) and the cell map shows the batteries all within 40mV of each other, than the pack is good.

Am I missing anything?

Sorry if I'm interupting the thread!
 
201 - 220 of 228 Posts
Top