Cell voltages during discharge - Log data wanted

[ScottyDont]

Hi all, I have been following along with the battery degradation discussions and was curious of the battery voltage during discharge. I was recently in cold weather and the engine didn't start properly, on the short drive to the petrol station I lost significant power but unfortunately wasn't logging. From another thread it looks like an improper recall, when I restarted the car the engine ran fine for the rest of the trip.

I have a discharge curve from an iMiev to compare to, allowing me to estimate the real capacity for my own car. The discharge curve compares well to the iMiev but I would like to get a few different battery logs to compare.

To log I used PHEV Watchdog and ticked both Logging Settings with an interval of 1 second.

It would be great if a few people could send me a log of a full discharge where occasionally they coasted in B0 for 5 seconds or so. I would especially be interested in a PHEV that has recently had a battery replacement to get a baseline of a perfect condition battery. If anyone is considering performing the new Lindqvist method or a dealer reset a log before and after would also be interesting. I haven't done it yet but I may use the fuel pump fuse trick to get a low SoC reading.

Be aware it may log GPS data, if you want more privacy just untick the GPS option before logging and PMing the file.

I have attached the comparison graph from my single log, for reference my car is showing 29.7Ah and has done 61,000km and 3.6L/100km

 

[elm70]

Discharge curve is very specific to:
- Temperature
- Load

In winter on a highway you might expect to see even 3.3v at 50% SOC

In summer on moderate usage, and after some regen .. 4.0v can be seen at same SOC

I don't see how you can use this information

Only valid information is Voltage per BMU SOC, after 2H rest .. but then you can see the BMU Voltage to SOC mapping curve per the BMU EEPROM

 

[ScottyDont]

Thanks elm70 for the feedback. I agree to both of those points and to minimise it I have done the following.

Load - I am only taking lightly loaded data points (coasting in B0). When you start coasting the voltage bounces back quickly then slowly increases from there. I am using a lightly loaded discharge curve from an iMiev so voltage should be closer to my sample data then a 2 hour rest.

Temperature - while loaded the voltage will sag more at lower temperature, but recovers similarly at light load. I expect the B0 coast voltage would recover to a reasonable value quickly but if it stays too low it can be identified through the battery temperature information in the log. While driving in EV mode the battery should warm up as well, though I haven't seen by how much in really low ambient temperatures.

 

[elm70]

Thanks elm70 for the feedback. I agree to both of those points and to minimise it I have done the following.

Load - I am only taking lightly loaded data points (coasting in B0). When you start coasting the voltage bounces back quickly then slowly increases from there. I am using a lightly loaded discharge curve from an iMiev so voltage should be closer to my sample data then a 2 hour rest.

Temperature - while loaded the voltage will sag more at lower temperature, but recovers similarly at light load. I expect the B0 coast voltage would recover to a reasonable value quickly but if it stays too low it can be identified through the battery temperature information in the log. While driving in EV mode the battery should warm up as well, though I haven't seen by how much in really low ambient temperatures.

I don't think you can even be accurate in this way

When voltage get depressed, due to high load and eventually low temperature, it can take hours before the voltage bounce back to the proper unload value .. this is why the BMU take care of the 2h rest before update the SOC after usage.

If you want a more consistent result ... we should make a log of the charging process, eventually after over 2h rest ... and ideally with warm temperature

But, this will require hours of logs ... something that only Anko was doing with his extra OBD2 equipment on the PHEV

 

[MadTechNutter]

I agree with elm70 that logging while driving is extremely variable and difficult to make conclusions.

I can not find original discharge curves for this battery to compare the voltages and DoD. SoC is not a useful parameter for these graphs as we need to see the voltage after how much energy we removed and not how much we might expect.

You might have noticed this link I posted in the Lyndqvist thread.
https://www.gs-yuasa.com/en/technic/vol9/pdf/009_01_026.pdf

In that report there are discharge charts for the LEV50 under different conditions, like load and temperature and it is reasonable to assume that the LEV40 follows very similar patterns.
In one chart you will see a graph for a 0.2C load. I put the selector in drive with handbrake on and switch on the headlights and wait a minute for the voltage to stabilise. This should give you roughly the 0.2C reading to compare with the chart. Probably good to print out the chart larger and draw in some lines to get more accurate readings.

It is interesting that your car's degradation is almost identical to mine. Despite that being rather disappointing I am starting to believe that 29Ah SoH after 60,000 is all we can expect from these batteries that are really quite small for the 2 ton vehicle they have to push around when they get fully recharged almost every day.
Maybe 2-3 Ah can be squeezed out from those magic resets with deeper discharge without harm but that is not very much.

Maybe some people who have reported up to 200,000km are long distance drivers where the petrol engine did all the work and the battery relatively rarely got cycled.

What kills batteries fast is daily fully charged overnight up to 4.1V but then used the next day and
strong regenerative charge B5 downhills over a minute each day.
Battery degradation all comes down to how these vehicles have been used.

 

[jaapv]

When you drive long distance the battery gets recharged all the time by the onboard generator...

 

[ScottyDont]

Thanks both for your input, good to be getting more ideas.

MadTechNutter, I had seen that pdf a while ago and originally used that for my voltages, but went with the iMiev discharge curve as it didn't contain the dip around 20 percent. I reread the document and it really suffers in low temperatures, I don't usually get <0C here so can't check voltage recovery myself. My logs show 11 degree C rise over the full drive, not as much as I expected. I added the curves from this document and could possibly scale voltage based on measured temperatures.

I agree reported SoC is not all that useful, but it allows me to compare different capacity cells, plus I have a scaling factor allowing me to shift the plot up and down as if the reported capacity was different. I have kept it at 1 for the initial attachment as it lined up ok, but my test down to 20% today is showing higher than expected voltages, more in line with what the SoC would be for a 33Ah pack.

I feel km driven is useless without long term fuel consumption, even then it only gives a slightly better picture of how much the battery has worked.

elm70, I wasn't aware you can log charging, I might have a look at that. The voltage at rest isn't what I am comparing, as the LEV50/LEV50N data is all constant discharge not rested values.

 

[MadTechNutter]

When you drive long distance the battery gets recharged all the time by the onboard generator...

...but not cycled.

It get charged to a certain level and basically stays there and not drained to 30% and back to 100% every 45km.

 

[MadTechNutter]

I agree reported SoC is not all that useful, but it allows me to compare different capacity cells, plus I have a scaling factor allowing me to shift the plot up and down as if the reported capacity was different. I have kept it at 1 for the initial attachment as it lined up ok, but my test down to 20% today is showing higher than expected voltages, more in line with what the SoC would be for a 33Ah pack.

Scotty, I don't quiet understand what you are doing.
How to you get the discharge values down to 20% ?
Are you driving ? ...or what load are you using?

 

[elm70]

When you drive long distance the battery gets recharged all the time by the onboard generator...

...but not cycled.

It get charged to a certain level and basically stays there and not drained to 30% and back to 100% every 45km.

No it is cycled.

It get charged up to 33/35% SOC and discharged down to 30%

At high speed the ICE will never stop, but still it charge and discharge the battery around the 30% SOC

Even at high speed, still it is needed the main battery for power the eMotor for reduce the drag, plus there is the 12v system that consume power and need to be charged by the main battery ... so the main battery it is all time used, and need to be charged all the time, unless the car run in pure EV mode

 

[MadTechNutter]

When you drive long distance the battery gets recharged all the time by the onboard generator...

...but not cycled.

It get charged to a certain level and basically stays there and not drained to 30% and back to 100% every 45km.

No it is cycled.

It get charged up to 33/35% SOC and discharged down to 30%

At high speed the ICE will never stop, but still it charge and discharge the battery around the 30% SOC

Even at high speed, still it is needed the main battery for power the eMotor for reduce the drag, plus there is the 12v system that consume power and need to be charged by the main battery ... so the main battery it is all time used, and need to be charged all the time, unless the car run in pure EV mode

My point is that the battery does not wear out under these conditions, it could last for decades like that!

I trying to prove that a battery that is used purely in EV mode, cycled everyday, can only last 60,000km when it's SoH drops below 80% and probably at 90,000 km it becomes so weak that it needs replacement.

....and that this whole argument why for some SoH lasts very long and for others it drops fast, is solely based on how they use it.

However I am very certain based on every other Li-IOn battery I have dealt with, that if the charge could be terminated around 90%, the battery used mostly in EV mode could last 120,000 km before it is 50% SoH and that discharging to 20% or even less would help to achieve that.
I am aware of the safety buffer required the keep the system active, mainly at cold temperatures, so probably not much can be done here with older batteries like mine.
If people with new cars were aware of this though, a 90% charge limit could save them a lot of grief later on.

 

[ScottyDont]

Scotty, I don't quiet understand what you are doing.
How to you get the discharge values down to 20% ?
Are you driving ? ...or what load are you using?

I removed the fuel pump fuse and drove it to 20% recording voltage at light load. The PHEV reduced power here to prevent further discharge of the battery, but if it is underreporting SoH then this will really occur at a higher point on the battery discharge curve. I added a 33Ah capacity calculation to show the lowest SoC was closer to 30% as if I got the BMU reset and it reported a higher SoH. The cell voltage curve will not change with as BMU reset, full is still 4.1V and empty is still ~3V.

 

[ScottyDont]

As to the cycling during hybrid mode, this causes less degradation at 30-35% SoC than if it is cycled from 95-100%, as the cell is under much less stress while discharged. The same reason that only charging to 90% provides cycle life increase >10% so if you can spare the range and inconvenience it's best to minimise your time near 100%.

 

[MadTechNutter]

I removed the fuel pump fuse and drove it to 20% recording voltage at light load. The PHEV reduced power here to prevent further discharge of the battery, but if it is underreporting SoH then this will really occur at a higher point on the battery discharge curve. I added a 33Ah capacity calculation to show the lowest SoC was closer to 30% as if I got the BMU reset and it reported a higher SoH. The cell voltage curve will not change with as BMU reset, full is still 4.1V and empty is still ~3V.

Ahhh a pure EV mode, naughty boy
I am a PHEV newbie, still have a lot of tricks to learn.

Maybe wire in a switch in series to the fuse connection for an EV Drag-mode(when fully charged) :mrgreen:
I have a lot of blank switch spaces because I got the simpleton version.

That doesn't harm anything(apart from the risk of depleting the battery)?
I could imagine that air gets in the fuel lines and the front electric motor keeps trying to start the engine, wearing out clutches or something?

 

[elm70]

I removed the fuel pump fuse and drove it to 20% recording voltage at light load. The PHEV reduced power here to prevent further discharge of the battery, but if it is underreporting SoH then this will really occur at a higher point on the battery discharge curve. I added a 33Ah capacity calculation to show the lowest SoC was closer to 30% as if I got the BMU reset and it reported a higher SoH. The cell voltage curve will not change with as BMU reset, full is still 4.1V and empty is still ~3V.

Ahhh a pure EV mode, naughty boy
I am a PHEV newbie, still have a lot of tricks to learn.

Maybe wire in a switch in series to the fuse connection for an EV Drag-mode(when fully charged) :mrgreen:
I have a lot of blank switch spaces because I got the simpleton version.

That doesn't harm anything? I could imagine that air gets in the fuel lines and the front electric motor keeps trying to start the engine, wearing out clutches or something?

About the fuel pump fuse

I did try it, and somehow it works .. but only somehow

It works fine on the first usage .. the car try only once to start the engine .. it fail .. it show an error on console and does not try anymore

But .. on next day or next restart .. it does try multiple time to start the ICE ... actually in a bad scenario it was trying to start the ICE all the time ... so I was afraid it could damage my engine.

So .. at the end .. I stop used it .. as well .. I notice an association between this fuse removal, and old fuel (if the car believe the fuel is over 3 months old, it will start the ICE immediately and for ever until new fuel is added) .. so that I had to remove the 12v battery for reset this

PS: The is no risk to over discharge the battery .. but there is a big risk to get stuck in trufic .. and it might need to connect the car to a charger before it will even try again to start the ICE ... so ... relative dangerous ... for explore the EV range .. it is mainly good for people that like to accelerate as much as possible without risking to start the ICE .. it really drive much more happy my PHEV without the fuse ... but ... due to the issues, even if i did acquire a 12v relay with a remote control for enable and disable the fuse automatically .. I did abandon this option (maybe there are multiple fuse to pull out, or I was pooling out the wrong one ... anyhow .. for my usage I don't see much gain in this ares)

 

[MadTechNutter]

I removed the fuel pump fuse and drove it to 20% recording voltage at light load. The PHEV reduced power here to prevent further discharge of the battery, but if it is underreporting SoH then this will really occur at a higher point on the battery discharge curve. I added a 33Ah capacity calculation to show the lowest SoC was closer to 30% as if I got the BMU reset and it reported a higher SoH. The cell voltage curve will not change with as BMU reset, full is still 4.1V and empty is still ~3V.

Ahhh a pure EV mode, naughty boy
I am a PHEV newbie, still have a lot of tricks to learn.

Maybe wire in a switch in series to the fuse connection for an EV Drag-mode(when fully charged) :mrgreen:
I have a lot of blank switch spaces because I got the simpleton version.

That doesn't harm anything? I could imagine that air gets in the fuel lines and the front electric motor keeps trying to start the engine, wearing out clutches or something?

About the fuel pump fuse

I did try it, and somehow it works .. but only somehow

It works fine on the first usage .. the car try only once to start the engine .. it fail .. it show an error on console and does not try anymore

But .. on next day or next restart .. it does try multiple time to start the ICE ... actually in a bad scenario it was trying to start the ICE all the time ... so I was afraid it could damage my engine.

So .. at the end .. I stop used it .. as well .. I notice an association between this fuse removal, and old fuel (if the car believe the fuel is over 3 months old, it will start the ICE immediately and for ever until new fuel is added) .. so that I had to remove the 12v battery for reset this

Thanks elm70, I thought that could happen, won't built a switch for the fuse wire.
But the fuel tank sensor, might get my attention one day.
Pretty sure it is not too hard to emulate an empty tank to disable the ICE and then an increase of 15l is automatically given when the sensor is connected again, assuming there actually is at least 15l in the tank.

 

[jaapv]

I trying to prove that a battery that is used purely in EV mode, cycled everyday, can only last 60,000km when it's SoH drops below 80% and probably at 90,000 km it becomes so weak that it needs replacement.

....and that this whole argument why for some SoH lasts very long and for others it drops fast, is solely based on how they use it.

However I am very certain based on every other Li-IOn battery I have dealt with, that if the charge could be terminated around 90%, the battery used mostly in EV mode could last 120,000 km before it is 50% SoH and that discharging to 20% or even less would help to achieve that.
I am aware of the safety buffer required the keep the system active, mainly at cold temperatures, so probably not much can be done here with older batteries like mine.
If people with new cars were aware of this though, a 90% charge limit could save them a lot of grief later on.

You forget that there is a wealth of empirical evidence. There are tens of thousands of older cars out there: in the Netherlands alone about 30.000 2013-2014 ones. None of these dire predictions have come to pass.
My car has been driven for 130.000 km and has been charged to "full" twice every working day for 5 years. I guess that of those 130.000 not more than 30.000 were uncharged motorway kms, the rest mainly EV Even then I usually drive the motorway with the charge button pressed. The battery is at about 80% SOH, the range at about 85%. That is a completely normal state for cars of that vintage.
The charge is indeed stopped automatically well below max charge by the charging algorithm.
It is not for nothing that Mitsubishi gives an 8 year warranty on the battery.

 

[MadTechNutter]

I trying to prove that a battery that is used purely in EV mode, cycled everyday, can only last 60,000km when it's SoH drops below 80% and probably at 90,000 km it becomes so weak that it needs replacement.

....and that this whole argument why for some SoH lasts very long and for others it drops fast, is solely based on how they use it.

However I am very certain based on every other Li-IOn battery I have dealt with, that if the charge could be terminated around 90%, the battery used mostly in EV mode could last 120,000 km before it is 50% SoH and that discharging to 20% or even less would help to achieve that.
I am aware of the safety buffer required the keep the system active, mainly at cold temperatures, so probably not much can be done here with older batteries like mine.
If people with new cars were aware of this though, a 90% charge limit could save them a lot of grief later on.

You forget that there is a wealth of empirical evidence. There are tens of thousands of older cars out there: in the Netherlands alone about 30.000 2013-2014 ones. None of these dire predictions have come to pass.
My car has been driven for 130.000 km and has been charged to "full" twice every working day for 5 years. I guess that of those 130.000 not more than 30.000 were uncharged motorway kms, the rest mainly EV Even then I usually drive the motorway with the charge button pressed. The battery is at about 80% SOH, the range at about 85%. That is a completely normal state for cars of that vintage.
The charge is indeed stopped automatically well below max charge by the charging algorithm.
It is not for nothing that Mitsubishi gives an 8 year warranty on the battery.

76% SoH at 60,000km in my case does not give me reason to believe that I will get anywhere near 130,000km.
...and I am reading similar cases from others, on this forum and elsewhere.

Maybe you mostly drove soon after a full charge so that it rarely stood around at 99% SoC.

No 8 year warranty for the battery in Australia.
Less than 1,500 (2014/15) were sold here but it looks like a lot of complaints of battery weakness come from Australia.
One might think it could have to do with the weather but It is not always hot here. In some areas it falls below 0˚.
Most of the year it is a comfortable 25˚ average, in the really hot areas there is a very low population density and only petrol heads live there.
In Europe where I lived for almost 30 years I can remember plenty of heatwaves too.

 

[jaapv]

Not really - my normal pattern of use is: unplug in the morning, drive 30-35 km to my work - with the odd detour for mail etc., a mix of urban, 80 kmh roads and motorway, plug it in at work, leave it until the afternoon, unplug, drive home, plug in and leave it overnight.
BTW, this week temperatures over here are between 35º and 40º. It does my range a world of good: yesterday I drove 38 km on 99% EV with the airco going full blast all the time.

 

[greendwarf]

Topography may be a bigger factor than temperature between Netherlands and Australia. Apart from the "mountains" in the South the Low Countries are relatively flat whereas there are lots of hills in the South East of Australia plus Great Dividing Range! As a result jaapav's commute is less likely to have the peaks of demand & regen - unless he is a "red-light racer" (which traffic density will discourage) - that might occur in similar daily driving in Australia (or even UK)? :idea: