Mach-E Extended Range Battery buffers and DCFC limits

[EVmodeler]

The plot shows the estimated open circuit voltage and the battery terminal voltage for 100 and 400 A current for both charge (+) and discharge (-).
The Voc curve model is estimated for Li-Ion (NMC?). The internal resistance Rint is also estimated based on scaling from similar cells. Rint is treated as approx constant, but would rise at high SOC for charging and also rise at low SOC for discharge. Rint would also drop as battery temperature rises. This plot is a simplified model to show trends and probable charge buffers and DCFC limits. If you have better data or estimates, please let me know

Charge Buffers
The charge buffer at the top is estimated to be small, about 6 Ah or a little less than 2% (absolute) SOC. Not much more energy can be stored here as the battery cell voltage starts to rise up more quickly.
The bottom charge buffer is larger, and estimated to be about 28 Ah or just less than 10% of Ah capacity. The cell voltage starts to drop very quickly ("falling off the cliff") below 10% SOC. Some energy (8 kWh) stored here, but difficult to access for discharge as the voltage drops. Pack cell uniformity and balance (and aging) come in to play also. Perhaps some additional capacity could be unlocked here, but not much likely (in my opinion). Note that as the pack voltage drops, the current required to produce a given power demand increases. So at low SOC, the pack power and thus motor power will be limited, which can change the acceleration response expected by the driver (so bad/dangerous).

The usable capacity would likely be displayed as "full" or 100% in the vehicle at 98% absolute SOC, and as empty or 0% near 10% absolute SOC. There is often a lot of confusion about these two different scales, one absolute for total capacity, and the other for usable capacity. Helpful to state which one you mean in discussions.

DC Fast Charge limits
The vehicle side DCFC components; wire, contactors (separate from main traction contactors?) and fusing in the vehicle are designed for some current limit. From the (estimated) properties of the battery, my estimate is that the current rating is 400 A, and this is near continuous rating for the possible ~40 min charging session. Most, but not all of that current/power goes into the battery; there are the DC-DC 12V loads while charging, and also thermal management (cooling and fans) since there is about 11 kW of losses generated in the battery at 400 A. The initial temperature of the pack at the start of DCFC as well as the ambient temperature for heat rejection also matter.

400 A current generates about 28V rise above the pack open circuit voltage, so as the SOC and Voc rise, eventually an upper voltage limit is reached (just like any constant current/constant voltage battery charging profile, including L2). If that voltage limit is 385V (4.1 V/cell), then the DCFC current and power would have to start to fall off at about 60% SOC. Because of the voltage rise during charging, the power at that point 385V * 400A = 154 kW, so a bit above 150 kW rating, but the true limit is almost certainly a current limit. The 150 kW power point would occur at 375V or around 48% SOC.

A bit more energy could be added to the battery at lower SOC and lower voltage, but that would require a bit higher current rating for not much return. 400 A cables and components are already pretty heavy and also pricey.

So - the 150 kW/400 A DCFC limit is mostly determined by the battery capability and properties, and are not at all likely to be increased (in my opinion).

400 A/288 Ah = ~1.4 C-rate continuous charge rate - pretty high for essentially steady state current.

The Standard range battery is 96S3P of the same cells (we assume), and a similar analysis can be carried out. With 3/4 the cells in parallel, the pack internal resistance is higher and the current limit is likely 300 A, while the pack voltage limits are slightly higher (due to two more cells in series). Hence the DCFC capability of 115 kW is due to the battery capability and properties appropriate for the vehicle, and not Ford choosing to eliminate 150 kW as an option (again, all in my opinion, but based on some knowledge and experience).

Peak Discharge for performance
Peak discharge for full acceleration from 258 kW AWD (both motors) requires at least 270 kW battery power which is 1000 A at 270V at around 50% SOC, but this discharge rate is for a very limited amount of time, probably less than 20 sec. (This 1000 A case is not shown on the plot.)
The main traction wiring may be rated for 400 A continuous, but be able to handle 2.5 times that (typical peak/cont in light duty vehicle traction systems) for short term transients just from the thermal mass of the cables and the current-squared-time (I2-t) capacity of the fusing.

As the SOC and battery voltage drop, the peak current limit will limit motor power as described above. When going to a drag race, start at a full charge - well known advice.

One other note, a fault or dead short across the battery terminals would result in current of about 4700 A, and the circuit protection (fusing) has to be able to clear (stop current flow) for this condition. But only once

Sponsored

 

[SnBGC]

I applaud the effort and appreciate your post. Didnt read past 1/3rd the way because I have no understanding of such things and I rarely DCFC anyway. Looks like some very detailed info that many others would find very interesting. Stuff like this is what keeps the automobile hobby alive.??

 

[EVmodeler]

I applaud the effort and appreciate your post. Didnt read past 1/3rd the way because I have no understanding of such things and I rarely DCFC anyway. Looks like some very detailed info that many others would find very interesting. Stuff like this is what keeps the automobile hobby alive.??

Well, yes it turned out a bit longer than I intended :/
There has been discussion of DCFC rate, and opening up more battery capacity, so I thought this may help understanding the hardware limits.

I also forgot to mention that the SAE J1772 standard for CCS has a 400 A current limit;
higher power requires higher battery voltage.

 

[sockmeister]

The plot shows the estimated open circuit voltage and the battery terminal voltage for 100 and 400 A current for both charge (+) and discharge (-).
The Voc curve model is estimated for Li-Ion (NMC?). The internal resistance Rint is also estimated based on scaling from similar cells. Rint is treated as approx constant, but would rise at high SOC for charging and also rise at low SOC for discharge. Rint would also drop as battery temperature rises. This plot is a simplified model to show trends and probable charge buffers and DCFC limits. If you have better data or estimates, please let me know

Charge Buffers
The charge buffer at the top is estimated to be small, about 6 Ah or a little less than 2% (absolute) SOC. Not much more energy can be stored here as the battery cell voltage starts to rise up more quickly.
The bottom charge buffer is larger, and estimated to be about 28 Ah or just less than 10% of Ah capacity. The cell voltage starts to drop very quickly ("falling off the cliff") below 10% SOC. Some energy (8 kWh) stored here, but difficult to access for discharge as the voltage drops. Pack cell uniformity and balance (and aging) come in to play also. Perhaps some additional capacity could be unlocked here, but not much likely (in my opinion). Note that as the pack voltage drops, the current required to produce a given power demand increases. So at low SOC, the pack power and thus motor power will be limited, which can change the acceleration response expected by the driver (so bad/dangerous).

The usable capacity would likely be displayed as "full" or 100% in the vehicle at 98% absolute SOC, and as empty or 0% near 10% absolute SOC. There is often a lot of confusion about these two different scales, one absolute for total capacity, and the other for usable capacity. Helpful to state which one you mean in discussions.

DC Fast Charge limits
The vehicle side DCFC components; wire, contactors (separate from main traction contactors?) and fusing in the vehicle are designed for some current limit. From the (estimated) properties of the battery, my estimate is that the current rating is 400 A, and this is near continuous rating for the possible ~40 min charging session. Most, but not all of that current/power goes into the battery; there are the DC-DC 12V loads while charging, and also thermal management (cooling and fans) since there is about 11 kW of losses generated in the battery at 400 A. The initial temperature of the pack at the start of DCFC as well as the ambient temperature for heat rejection also matter.

400 A current generates about 28V rise above the pack open circuit voltage, so as the SOC and Voc rise, eventually an upper voltage limit is reached (just like any constant current/constant voltage battery charging profile, including L2). If that voltage limit is 385V (4.1 V/cell), then the DCFC current and power would have to start to fall off at about 60% SOC. Because of the voltage rise during charging, the power at that point 385V * 400A = 154 kW, so a bit above 150 kW rating, but the true limit is almost certainly a current limit. The 150 kW power point would occur at 375V or around 48% SOC.

A bit more energy could be added to the battery at lower SOC and lower voltage, but that would require a bit higher current rating for not much return. 400 A cables and components are already pretty heavy and also pricey.

So - the 150 kW/400 A DCFC limit is mostly determined by the battery capability and properties, and are not at all likely to be increased (in my opinion).

400 A/288 Ah = ~1.4 C-rate continuous charge rate - pretty high for essentially steady state current.

The Standard range battery is 96S3P of the same cells (we assume), and a similar analysis can be carried out. With 3/4 the cells in parallel, the pack internal resistance is higher and the current limit is likely 300 A, while the pack voltage limits are slightly higher (due to two more cells in series). Hence the DCFC capability of 115 kW is due to the battery capability and properties appropriate for the vehicle, and not Ford choosing to eliminate 150 kW as an option (again, all in my opinion, but based on some knowledge and experience).

Peak Discharge for performance
Peak discharge for full acceleration from 258 kW AWD (both motors) requires at least 270 kW battery power which is 1000 A at 270V at around 50% SOC, but this discharge rate is for a very limited amount of time, probably less than 20 sec. (This 1000 A case is not shown on the plot.)
The main traction wiring may be rated for 400 A continuous, but be able to handle 2.5 times that (typical peak/cont in light duty vehicle traction systems) for short term transients just from the thermal mass of the cables and the current-squared-time (I2-t) capacity of the fusing.

As the SOC and battery voltage drop, the peak current limit will limit motor power as described above. When going to a drag race, start at a full charge - well known advice.

One other note, a fault or dead short across the battery terminals would result in current of about 4700 A, and the circuit protection (fusing) has to be able to clear (stop current flow) for this condition. But only once

This was all REALLY interesting to me. I don't know why I never saw this post before. Thank you for this very detailed breakdown.

What are your thoughts as to what is going on with the battery to limit the usable capacity to 88kwh? Is a 100% state of charge at 88kwh still actually only at 98% SoC -- given the very large buffer?

Can the usable portion be easily increased without a hardware limitation?

 

[EV Lab]

This was all REALLY interesting to me. I don't know why I never saw this post before. Thank you for this very detailed breakdown.

What are your thoughts as to what is going on with the battery to limit the usable capacity to 88kwh? Is a 100% state of charge at 88kwh still actually only at 98% SoC -- given the very large buffer?

Can the usable portion be easily increased without a hardware limitation?

The capacity limits and reasons are already described in the analysis above... 100% indicated at 98% SoC and 0% indicated at ~10% SoC. Using those areas would require a much more aggressive BMS or accepting greater risk of increased aging or reduced lifespan.

 

[Billyk24]

A temperature factor is being ignored. Try again with very cold, moderate and very hot temperatures.

 

[machefan]

Great post!

Summary from the MME manual for the average Joe

 

[Billyk24]

Great post!

Summary from the MME manual for the average Joe

Allowing your battery to cool...... (more so during hot summer weather) There is another thread in which a question has been asked on what is the battery temperature during driving, during L2 and DCFC charging. As on now, a couple owners are trying to find the PIDS to read such data.

 

[sockmeister]

The capacity limits and reasons are already described in the analysis above... 100% indicated at 98% SoC and 0% indicated at ~10% SoC. Using those areas would require a much more aggressive BMS or accepting greater risk of increased aging or reduced lifespan.

I didn't realize that was referring to the 10kWh capacity buffer -- I thought it was about the battery's current limitations assuming it's locked at 88kwh.
So in the end, you're saying it's unlikely much additional capacity will be unlocked? Why do other manufacturers have much smaller buffers for the same technology?

 

[sockmeister]

@EVmodeler also - where did you get this data from?

 

[EVmodeler]

A temperature factor is being ignored. Try again with very cold, moderate and very hot temperatures.

Its an early, approximate model with the assumptions stated, as an example of how the buffers and voltage/current limits come into play.

Do you have detailed info of internal resistance vs temperature and SOC?
Please share.

 

[Billyk24]

Do you have detailed info of internal resistance vs temperature and SOC?

No. It is well known that too cold and too hot will effect the battery capabilities.

 

[EVmodeler]

@EVmodeler also - where did you get this data from?

The plot is based on reasonable assumptions from other (mostly) similar batteries. The purpose is to illustrate where the buffers and DCFC limits come from - approximately. As we gather more data specific to the MME, the model can be improved, and perhaps applied to low and high battery temperatures.

 

[RonTCat]

I didn't realize that was referring to the 10kWh capacity buffer -- I thought it was about the battery's current limitations assuming it's locked at 88kwh.
So in the end, you're saying it's unlikely much additional capacity will be unlocked? Why do other manufacturers have much smaller buffers for the same technology?

Either...
- They just go for max EPA range, and burn down their battery fast, or
- People are just guessing what the buffers are, and have no idea.

 

[EVmodeler]

Do you have detailed info of internal resistance vs temperature and SOC?

No. It is well known that too cold and too hot will effect the battery capabilities.

Right. So this example is for moderate temperatures as a starting place.

As it says right at the top of the post:
"The Voc curve model is estimated for Li-Ion (NMC?). The internal resistance Rint is also estimated based on scaling from similar cells. Rint is treated as approx constant, but would rise at high SOC for charging and also rise at low SOC for discharge. Rint would also drop as battery temperature rises. This plot is a simplified model to show trends and probable charge buffers and DCFC limits."

Sponsored