Hyundai IONIQ Forum banner

1 - 20 of 36 Posts

·
Registered
Joined
·
132 Posts
Discussion Starter #1 (Edited)
Reading through this forum I find in many posts a statement (a belief?) that the Ioniq EV battery has total buffer (margin) of about 10%. I could not find any source to this claim.

Looking at the EPA test results (136 mpge and 124 miles AER) we can calculate:
33.7/136*124=30.73 kWh (33.7 kWh is equivalent to 1 gal. benzine according to EPA).
These ~30.73 kWh consumed to fill-up back the battery is (as always with EPA) consumption from the wall, i.e. payable electricity.

To calculate battery usable capacity we have to take into account charging losses. I don't know if the charging back was done at 120V (L1) or 240V (L2). In early years of smaller battery PHEVs it was done by EPA via L1 but batteries these days are much larger.

Anyways, from an objective and reputable source like EPA, assuming 12-16% charging losses, we can calculate the battery usable capacity to be something in between 25.8 kWh to 27 kWh. Not 28 kWh as some assumed/suggested.

If we take 28 kWh as battery total capacity (or size) as indicated in all Hyundai publications that I have seen, we get a (total) buffer size of something in between 3.4% to 7.8% and not ~10%!
Battery total capacity may be (despite publications) larger than 28 kWh? maybe, but by how much? we don't know so why guessing 10% margin?
Personally I believe in Hyundai publications which means the total safety margin is very slim, in the order of 3.4-7.8%.
 

·
Super Moderator
Joined
·
1,584 Posts
In contrast to most other brands, publications of battery capacity by Hyundai always refer to usable capacity, not total capacity. For the overall capacity it has been told (but not written by Hyundai) that it is around 31 kWh.
 
  • Like
Reactions: JejuSoul and giora

·
Registered
Joined
·
132 Posts
Discussion Starter #3
O.K.
If it is always referred to as usable size then I tend to accept that the charge in the EPA test was done at 240 V (L2) with charging losses of 10% (a bit optimistic to my taste) giving usable capacity of 27.7 kWh rounded up to 28.
I assume the concepts of ~31 kWh total and 10% buffer go together, one leading to the other and still only a theory without a proof. So is the theory of "eating the upper buffer first", time will tell if not Hyundai.
 

·
Registered
Joined
·
424 Posts
Take a look at my plots (this and this). We know usable capacity is ~28kWh, it also must drop a bit if you discharge faster. The faster you discharge, the less energy the battery yields.

We know the BMS DC voltimeter and amperimeter are correct, because they match what is told by DC fast chargers (discounting cable losses). We know the BMS Cumulative Energy Charged counters are correct, because they match the area underneath the power curve (measured by numerical integration). The Cumulative Energy Discharge counters must be correct... and match the consumption displayed by the on-board computer.

We know that voltage at 100% is about 4.14V at rest. Li-Ion is usually charged to 4.2V and in some cases, stretched to 4.3V and beyond that. So you could charge it more. We know voltage at 1% is ... (can't remember, but above 3.2V)... Li-Ion can usually be discharged to a lower voltage than that...

The EPA number measures what the amount of energy that the BMS lets EPA charge. It's the usable capacity.

Another proof of the buffer: Li-Ion batteries degrade. It's unavoidable. However, after one year and 42000km, my battery still yields 27.*kWh and the BMS reports 100% SoH. Just as new. The only way that is possible, is a buffer.
 

·
Registered
Joined
·
132 Posts
Discussion Starter #5
...The EPA number measures what the amount of energy that the BMS lets EPA charge. It's the usable capacity...
What I said, no debate here. However the number shown by them includes charging losses i.e. kWh from wall and to find the usable capacity you must deduct the charging loss.
 

·
Registered
Joined
·
41 Posts
Take a look at my plots (this and this). We know usable capacity is ~28kWh, it also must drop a bit if you discharge faster. The faster you discharge, the less energy the battery yields.

We know the BMS DC voltimeter and amperimeter are correct, because they match what is told by DC fast chargers (discounting cable losses). We know the BMS Cumulative Energy Charged counters are correct, because they match the area underneath the power curve (measured by numerical integration). The Cumulative Energy Discharge counters must be correct... and match the consumption displayed by the on-board computer.

We know that voltage at 100% is about 4.14V at rest. Li-Ion is usually charged to 4.2V and in some cases, stretched to 4.3V and beyond that. So you could charge it more. We know voltage at 1% is ... (can't remember, but above 3.2V)... Li-Ion can usually be discharged to a lower voltage than that...

The EPA number measures what the amount of energy that the BMS lets EPA charge. It's the usable capacity.

Another proof of the buffer: Li-Ion batteries degrade. It's unavoidable. However, after one year and 42000km, my battery still yields 27.*kWh and the BMS reports 100% SoH. Just as new. The only way that is possible, is a buffer.

May I ask what your charging habits are (do you try and discharge close to 20% and charge up to 80%?) Having only had my ioniq EV a week, I am curious about the whole battery degradation thing. I do not drive as much as you (42000km in one year). I put on about 15000km a year, so i plan on owning this my2019 for about 10 years.
Thank you
Lester
 

·
Registered
Joined
·
424 Posts
What I said, no debate here. However the number shown by them includes charging losses i.e. kWh from wall and to find the usable capacity you must deduct the charging loss.
Sure, and we can measure all that. The CEC and CED counters tell so much! My point was that we can trust these counters. Every measurement device has error, but they aren't intentionally biased...
They count energy already in DC charging or discharging the battery. So, if you have an energy meter at the EVSE (counting AC energy), you can know how much energy was lost, grossly, in the AC to DC conversion.

Then, the energy in DC that charges the battery from 10% to 100% isn't the same yielded by discharging the battery from 100% to 10% (which is the useful energy). Li-Ion batteries have 99% coulombic efficiency, meaning you recover all the charge, in Ah, that you put in there. However, because voltage rises during charging, and lowers during discharging, due to internal resistence and hysteresis, you don't get the same kWh (you get, say, 94% in practice). The remaining energy is lost as heat in the battery.

Whenever you charge x% of the SoC by an amount of y kWh (measured by CEC), you can estimate the energy needed to charge from 0% to 100% by dividing y kWh by x%. Whenever you discharge x% an amount of y kWh (measured by CED - CEC), you can estimate the energy yielded by discharge from 100% to 0%... And whenever you take note of these numbers, take note of the charging or driving time, because the amounts of energy vary with charge/discharge speed (the faster, the more heat loss in the battery). What you get is typically about 27.6kWh or better useful capacity for a 5h drive (0.2C discharge) or 27.2, even 26.9kWh useful capacity for a 1h30' drive (0.67C discharge, motorway). For charging, about 29kWh normal charge, and 30kWh fast charge.

So, we can measure the loss from the EVSE to the DC poles of the battery, then the heat loss in the battery.... If we manage to decode the OBD data from the OBC, then we can be even more picky and separate the loss in the cable from the EVSE and the loss in the AC/DC conversion. But it wouldn't be my priority.

Here's a charging session where I varied the commanded current that the EVSE communicates to the car. I have an energy meter (PZEM) at my flat's entrance, measuring total "House power". A 40m wire then goes down to the garage (4 story building). Then the EVSE measuring the energy sucked by the car "AprEVSE power". Then the DC voltage times current (power at the poles of the battery). With that setup, I can easily estimate the loss in the wire to the garage. To check, "Remainder" is the power to the remainder of the house, where the oscillation of the refrigerator is visible. "Apr commanded" is the available current communicated by the EVSE to the car. Apr is the nick of the guy who assembled my EVSE (this is an international acknowledgement).



Because I varied the commanded current, I can now plot OBC and wire losses as a function of the commanded current. Note that there is some noise because it was only left about 5-10 minutes in each stage...



So, I can plot the charging efficiency of the OBC, the combined efficiency of the OBC and the wire (it hasn't been easy to replace that wire for a thicker one). And the heat loss in the battery is not in these plots (but it was in the other posts).



And sorry about the chart titles in portuguese...

Now, about degradation...

I understand your car is a lot newer than mine, right? At least you're newer in the forum...
We can measure it. We can compare our two cars.

We have to do two things. Er... three things:
  • We have to plot SoC BMS as a function of Displayed SoC. This should be a straight line. Otherwise, whenever you are below 30% or at 100% just tell me the values.
  • We have to plot CEC kWh increase as a function of Displayed SoC during a long charging session (say, 16A is more convenient to me). This should be a straight line, to the best abilities of the BMS (he'll try as hard as he can).
  • We have to plot DC voltage during that same long charging session at controlled speed. This won't be a straight line. Usually, we want open-circuit voltage at rest, meaning no current flowing in or out for a few hours. But I guess, a charging session at controlled speed (16A) may be easier to replicate.
We want to know if my charge charges to a higher voltage than yours, or if it discharges deeper than yours, to make up for loss of capacity. We need the second plot to know if usable capacity changed or not. We need the first plot to test the idea that the buffer capacity corresponds to what the BMS SoC sugests.

What do you say?
 

·
Registered
Joined
·
424 Posts
May I ask what your charging habits are (do you try and discharge close to 20% and charge up to 80%?)
I'd like to answer that on this other thread. For several days, I've been meaning to post on that thread, but I haven't managed to find the time. I'm not super careful. My Ioniq has to endure a 12km motorway bridge daily, cruised between 110km/h and 130km/h, very windy (near the sea), plus, I do a lot of 300km motorway trips on weekends, with a DCQC in the middle, frequently. You have to use the car.
 

·
Registered
Joined
·
132 Posts
Discussion Starter #9 (Edited)
...What do you say?
I say I am extremely impressed with the amount and deepness of the research work you are doing.
I do not own an Ioniq EV yet, in fact it was not yet launched here. Some more details in this post
The concept of using the upper buffer first as battery degrades (as reported in several posts here) is new to me and very intriguing (if proved to be the designed strategy):
I can see the advantages in the short run, but is it good for the long run? (years of having no upper buffer).
Will it not create situations where the battery is over-charged?
Say you have no upper buffer anymore and car is kept fully charged for hours, in which temp drops 10-15 degC or even more so battery may be in over-charged state for the lower temp?
"Drive immediately or close to finish 100% charging" is of course a very good advice but not always practical or kept.
 

·
Registered
Joined
·
424 Posts
@migle, I am in awe over your expertise on this topic. Thank you for the enlightening posts!
Nah! No expertise. This is what you pick in these forums over time. Here, also NissanLeafPT.com, and then some reading at batteryuniversity.com.
I admit to being handy with gnuplot, which is a great graphics package, particularly for many plots...
And yes, plots are always enlightening...

I do not own an Ioniq EV yet, in fact it was not yet launched here.
Oh! Ok.... Anyway, if anyone else wants to collect data, we could perhaps dive deeper into this issue....
Otherwise, maybe I can already compare my battery with my own battery a few months ago...

Say you have no upper buffer anymore and car is kept fully charged for hours, in which temp drops 10-15 degC or even more so battery may be in over-charged state for the lower temp?
I think you're taking the expression "you have no upper buffer anymore" too literally. To put it simply, naturally, there may be another buffer on top of the first buffer... The first one was depleeted, and the second will be kept forever... The first exists to hide loss of capacity, yes. The second exists for safety.

Then, I don't think we know for sure how much buffer there is at top and at bottom. Sure, discharging deeper is also bad, because the heat loss increases... Anyway, until we see something meaningful in the voltages, we really don't know much. After two years, we still know very little about loss of capacity in Ioniq, and that alone is a curious fact. We know there is a member in the german forum who has a slightly degraded batttery, but it only charges at a 150kW fast charger, sometimes several times a day, and it did over 125000km...

I think Hyundai did a tremendous job in designing a quality product. These EVs really stand out. Some other EVs have the quality of the interiors as their main selling point. These stand out because of their great performance.

Would I have preferred to enjoy the full capacity of my battery when it was new, rather than having it hidden for latter? I don't know... I think psychologically it works very well. I like the feeling of looking at the SoH and it's still at 100% and noticing no difference in capacity.
 

·
Registered
Joined
·
424 Posts
Anyway, if anyone else wants to collect data, we could perhaps dive deeper into this issue....
If anyone wants to compare his battery with mine, I would still be interesting. If your plotting skills need brushing, I can plot the data.
Particularly interesting would be: a brand new car, recently imported; or a very used car, with lots of fast charges.

For this, I think we should create a standard testing procedure. One that any Ioniq user can easily reproduce. I propose this, for which I request your comments.

  • Because measuring OCV at rest is very hard to do in a controlled fashion, the second best clearly is a normal charging session using the EVSE provided with the vehicle (called ICCB in the manual, a.k.a. soap), regulated for High level, which should be 12A.
  • We would only concern ourselves with a charging session at 230V AC, inside a garagem, before winter, and on a normal day (not after a trip, for instance).
  • Torque Pro would have to be used during the session, configured to log to a file, using the options "Synchronous Logging" and "Format log values". GPS and G-sensors undesirable. It should be configured to log: Auxiliary Battery Voltage, Battery Current, Battery DC Voltage, Cumulative Charge Current, Cumulative Discharge Current, Cumulative Energy Charged, Cumulative Energy Discharged, Maximum Cell Voltage, Minimum Cell Voltage, State of Charge BMS and State of Charge Display.
  • It would have to charge to 100%, preferrably since under 20%, the deeper, the better.
  • Additionally, if possible, 15 minutes after the charge ended, just turn on the car, let Torque Pro record voltage, then turn off. The same 1h after charging.
  • This should be enough to plot SoC BMS over SoC display, then CCC increase over SoC display, then CEC over SoC display and then Battery DC Voltage over SoC Display and report the Battery DC Voltage at rest some time after charging.
Anyone has such patience?
 

·
Registered
Joined
·
23 Posts
This looks like a useful thread to resurrect for the 38 kWh Ioniq. Anyone got ideas on the usable capacity and safety margins for this new model?

Personally, I'm a little worried that my battery has already got a bit of degradation as an ex-demo. I seem to have access to around 30 kWh of the battery with just 7000 miles on the clock.

E.g. 130 miles at 4.8 miles / kWh discharges from 100% to 8%.

130 ÷ 4.8 = 27 kWh used
Adjust for 8% remaining:
27 ÷ 0.92 = 29.4 kWh full discharge

Any other experiences to share?
 

·
Registered
Joined
·
190 Posts
E.g. 130 miles at 4.8 miles / kWh discharges from 100% to 8%.

130 ÷ 4.8 = 27 kWh used
Adjust for 8% remaining:
27 ÷ 0.92 = 29.4 kWh full discharge
Well you with 28kWh Ioniq that is in full health you typically get 26,5kWh out of the battery judging from the screen (miles travelled * kWh spent/mile). If you use this on your car it would be 36,1kWh on the screen. It could very well be that you had massive heat losses or the car has a different way of hiding the buffers up and down. This would mean that you could drive the car below 0% available on the screen.
 

·
Registered
Joined
·
784 Posts
This looks like a useful thread to resurrect for the 38 kWh Ioniq. Anyone got ideas on the usable capacity and safety margins for this new model?

Personally, I'm a little worried that my battery has already got a bit of degradation as an ex-demo. I seem to have access to around 30 kWh of the battery with just 7000 miles on the clock.

E.g. 130 miles at 4.8 miles / kWh discharges from 100% to 8%.

130 ÷ 4.8 = 27 kWh used
Adjust for 8% remaining:
27 ÷ 0.92 = 29.4 kWh full discharge

Any other experiences to share?
I don't see how the method you have used to calculate battery capacity can be right as the miles/kWh you achieve depends on driving style and whether heater/AC is used.
A more accurate method to determin battery SOH is to use an OBD11 dongle and smart phone app to read the individual cell voltages at 100% charge. There are some threads on here that explain how to do that in more detail.
 

·
Registered
Joined
·
23 Posts
I don't see how the method you have used to calculate battery capacity can be right as the miles/kWh you achieve depends on driving style and whether heater/AC is used.
A more accurate method to determin battery SOH is to use an OBD11 dongle and smart phone app to read the individual cell voltages at 100% charge. There are some threads on here that explain how to do that in more detail.
Is it not a good first-pass calculation though? I know the car charge % at the beginning and end of the journey. And I know the reported miles / kWh specifically for that journey.
 

·
Super Moderator
Joined
·
1,584 Posts
I think the numbers are too extreme to ignore and this cannot be due to degradation. Maybe you should ask your dealer to measure if there are a number of battery cells out of order. If so they should be replaced for free.

Another option is that your efficiency or SoC indication is out of order. If you count 0.92*38.3 = 35.2 kWh and you have used that for 130 miles, it would be 3.7 miles/kWh, which is really low. Then I would expect you to drive at a very high speed. Or a storm?

Or is it not one trip but several trips? Each with a new AC investment? Or did you stand still for a long time with the car on?

The buffer of the 38 kWh battery is not precisely known, but it was guessed that it is at most 1.5 to 2 kWh, so less than for the 28 kWh.
 
  • Like
Reactions: Knifyspoony

·
Super Moderator
Joined
·
1,584 Posts
By the way, in your post in another thread you show other numbers:

I did the return journey of my previously posted one, following the wise words of sensei Chewy. Achieved 5.9 mi/kWh on 130 mile journey.
I have another 10% efficiency to go to reach Chewy's numbers!
Picture attached for the full info, it was 28% battery when I stopped. Started at 100%.
Generally I found it better to turn off regen on the motorway, and I found massive gains from following large vehicles going less than 65mph such as lorries!
 

·
Registered
Joined
·
190 Posts
Or is it not one trip but several trips? Each with a new AC investment? Or did you stand still for a long time with the car on?
I'm pretty sure the AC and heating are included in the consumption displayed on the screen. Heat losses before the motor control unit and converters are probably not included in the displayed consumption. It is easy to check that the AC is included, just sit in your car for a couple of minutes before starting your drive and the consumption number will be in huge excess compared to the energy required for driving.
 

·
Registered
Joined
·
23 Posts
I think the numbers are too extreme to ignore and this cannot be due to degradation. Maybe you should ask your dealer to measure if there are a number of battery cells out of order. If so they should be replaced for free.

Another option is that your efficiency or SoC indication is out of order. If you count 0.92*38.3 = 35.2 kWh and you have used that for 130 miles, it would be 3.7 miles/kWh, which is really low. Then I would expect you to drive at a very high speed. Or a storm?

Or is it not one trip but several trips? Each with a new AC investment? Or did you stand still for a long time with the car on?

The buffer of the 38 kWh battery is not precisely known, but it was guessed that it is at most 1.5 to 2 kWh, so less than for the 28 kWh.
The example of 130 miles at 4.8 mi/kWh was a single journey on a single charge. Mixed driving with some motorway and some slower roads in 13 deg C dry weather (IIRC) That efficiency number seems quite correct comparing to others driving in similar conditions.

The other numbers I had was for the return leg of that, again 130 miles on a single charge but I chose to drive very economically and achieved 5.9 mi/kWh. With external ambient of around 18 deg C and dry weather.

The sums for this trip show:

E.g. 130 miles at 5.9 miles / kWh discharges from 100% to 28%.

130 ÷ 5.9 = 22 kWh used
Adjust for 28% remaining:
22 ÷ 0.72 = 30.5 kWh full discharge

Compared to 29.4 kWh for full discharge on the other trip. So my numbers are fairly repeatable within a suitable error margin for the efficiency only being to 2 significant figures.

I will contact Hyundai and maybe get my first annual check done before it's due in September.

Do you really think they'll replace bad cells? It's well within the 70% capacity of the battery warranty.
 
1 - 20 of 36 Posts
Top