Running the Battery to 0%

[Box Cat]

You can have the data he uses to arrive at his conclusions. See post 46. He’s trying to be as helpful as he can within the constraints of legal agreements of his employer. Anyone is free to disagree / disregard the advice given.

Thank you for clarifying, I must have missed the proprietary constraints …

And don’t get me wrong, I definitely appreciate his and everyone’s posts.

As for the data, given the terms, the ROI for getting it to make sense of the advice is close to zero .

Sponsored

 

[Grumpy2]

The following is taken from an older report of a HPPC test of a pouch similar to ours, but with a different cathode chemistry. We must wait for results of our pouch cells to know if it applies, however it is of interest if you regularly charge to 100% without needing to.

Search for: Characterization of High-Power Lithium-Ion Cells During Constant Current Cycling Part I. Cycle Performance and Electrochemical Diagnostics
Joongpyo Shim and Kathryn A. Striebel*

The report found an equal amount of added resistance within the cells after charging to 100% capacity 240 times; as is found when charging 70% capacity 1000 times. In addition, they determined the loss of capcity is not linear over the first 240 cycles, but the additional internal cell resistance can be measured before half that number of cycles.

 

[Yellow Buddy]

The report found an equal amount of added resistance within the cells after charging to 100% capacity 240 times; as is found when charging 70% capacity 1000 times.

Great, now I’m sitting here going. Let’s see…1, 2, I think that was 3. 4…

 

[Box Cat]

Interesting, although older is quite older ... (2003) , thanks for sharing!
So, at 0.5C, room temperature, 1000 cycles, they report 14% capacity loss for 70% DOD, and 70% capacity loss for 100% DOD, under the HPPC test.
They do say at the end that: "The large capacity and power losses in this cell chemistry come mainly from increases in the bulk and/or interfacial impedance of the cathode".
As you say, I would expect the NMC9 to behave quite differently.

 

[Box Cat]

This is a 5 year old article announcing the SK NMC9: NCM 90: successor of NCM 811 battery cells - Posts - PushEVs, including some comparative data pasted below:

An excerpt regarding the NMC9:

The NCM 90 cathode when cycled between 2,7 and 4,3 V at 0,5 C retains 90 % of its initial battery capacity after 100 cycles. Therefore, a new electric car with a 600 km range would have a 540 km range after 57.000 km [(600 + 540) / 2 x 100)] in these conditions. The good news is that no electric car battery is cycled in these extreme conditions. They are far better protected.

I would like to see how the NCM 90 cathode behaves when cycled between 2,8 and 4,1 V, it’s reasonable to expect that the cycle life is much better than only 90 % capacity retention after 100 cycles. It has to be good enough to allow a standard battery warranty, which for most EVs guarantees a minimum capacity of 70 % for eight years or 160.000 km (100.000 miles).

Does the last paragraph sound familiar?

 

[MickeyAO]

This is a 5 year old article announcing the SK NMC9: NCM 90: successor of NCM 811 battery cells - Posts - PushEVs, including some comparative data pasted below:

An excerpt regarding the NMC9:

The NCM 90 cathode when cycled between 2,7 and 4,3 V at 0,5 C retains 90 % of its initial battery capacity after 100 cycles. Therefore, a new electric car with a 600 km range would have a 540 km range after 57.000 km [(600 + 540) / 2 x 100)] in these conditions. The good news is that no electric car battery is cycled in these extreme conditions. They are far better protected.

I would like to see how the NCM 90 cathode behaves when cycled between 2,8 and 4,1 V, it’s reasonable to expect that the cycle life is much better than only 90 % capacity retention after 100 cycles. It has to be good enough to allow a standard battery warranty, which for most EVs guarantees a minimum capacity of 70 % for eight years or 160.000 km (100.000 miles).

Does the last paragraph sound familiar?

Something seems off in these graphs. Why are they overcharging the cell? Vmax (maximum voltage) should be 4.2 V, but the graphs show going above this. When I program my cyclers, the safety limits are set to 4.25 V, and they will shut down if the voltage goes over this value. Did they have an agenda to show how bad NMC 9 1/2 1/2 would perform? Also, it's lazy to label this chemistry as 90 instead of including the half for Manganese and Colbolt.

 

[Box Cat]

Well, they go as high as 4.5V which is even worse.
When I purchase a cell I get the specs from the manufacturer and set the charger/ BMS accordingly.
I assume 4.5V is within the safe range for testing the cell.

Yeah, people are lazy, but it is probably ok in the context. Also the article does not provide the whole paper so details may be missing.

 

[Box Cat]

Most probably, since they call it NCM 90 the CM portion might not have been disclosed at that time.

 

[DesertEV]

Not selling...just don't like the HPPC (Google it if you want) test below 20% SOC We test the cell directly at actual cell capacity (Vmax to Vmin static capacity before HPPC test).

Even worse above or below 25C.

We have finished characterization testing and are moving into calendar and cycle life testing. Pilot run starts by Friday

So based on your data are you saying it is recommended to be in the 20% to 70% range? I only went below 20 once to 15% and only went to 100% once for a trip. Any concerns there or just minimize those as much as possible? Thanks

 

[Box Cat]

I googled the subject a few days ago and read this department of energy report of 2015: Battery Test Manual For Electric Vehicles, Revision 3 (Technical Report) | OSTI.GOV, explaining the procedures of several EV battery tests, including HPPC.

In page 8, it says HPPC tests, after 10% increments, terminate at Vmin0 (whatever the manufacturer will tell you what a safe Vmin0 is), which seems what SK has done in those tests comparing the three different cathode compositions (you were right @MickeyAO , they were just lazy with NMC 90. It was spelled out to 9 .5 .5 somewhere in the article by the author who must have had access to the data ...).

What would be wrong going down to Vmin0 in an HPPC test?
Wouldn't you want to characterize the full range?

For temperature and other discharge current rates and other parameters, I am sure SK will have done far more than those charts show, which may even be outdated.

One thing that I do not understand is that in these posts we seem to imply a direct correlation between cell SoC tests and truck SoC (displayed on the console).

But these can be rather different stories since the Lighting has a 10% buffer (144-131)/131 ~ 10% (referred to the 131kWh that is published and visible to the customer) with lots of cells arranged in modules in a battery pack, temperature management, BMS, etc.

Ford is free to use whatever Vmin, Vmax, xC limits, etc. that make more sense to limit degradation over a broad range of usage patterns which eventually results in the catch-all 70% loss after 8y/100k miles warranty.

The truck has been out for years, and I assume all relevant tests will have been done before trucks are sold.

Am I missing anything?

Also, @MickeyAO , if you don't mind me asking, (and this is just my profound ignorance) what (or why) is being tested now that was not tested before by SK/Ford?

 

[MickeyAO]

I googled the subject a few days ago and read this department of energy report of 2015: Battery Test Manual For Electric Vehicles, Revision 3 (Technical Report) | OSTI.GOV, explaining the procedures of several EV battery tests, including HPPC.

In page 8, it says HPPC tests, after 10% increments, terminate at Vmin0 (whatever the manufacturer will tell you what a safe Vmin0 is), which seems what SK has done in those tests comparing the three different cathode compositions (you were right @MickeyAO , they were just lazy with NMC 90. It was spelled out to 9 .5 .5 somewhere in the article by the author who must have had access to the data ...).

What would be wrong going down to Vmin0 in an HPPC test?
Wouldn't you want to characterize the full range?

We actually do a 95% SOC and a 5% SOC pulse in our HPPC tests just to help calibrate the settings for our cycle life testing. The 0% SOC step is to find the voltage curve to Vmin. A pulse at 0% SOC would drive the cell below Vmin, so you don't want or need that pulse, nor 100% SOC in the HPPC test.

For temperature and other discharge current rates and other parameters, I am sure SK will have done far more than those charts show, which may even be outdated.

One thing that I do not understand is that in these posts we seem to imply a direct correlation between cell SoC tests and truck SoC (displayed on the console).

I have static capacity results for different C rates and temperatures, but nothing really stands out as concerning.

You will notice that I specifically state I don't want to go below actual 20% SOC due to what I saw.

But these can be rather different stories since the Lighting has a 10% buffer (144-131)/131 ~ 10% (referred to the 131kWh that is published and visible to the customer) with lots of cells arranged in modules in a battery pack, temperature management, BMS, etc.

Ford is free to use whatever Vmin, Vmax, xC limits, etc. that make more sense to limit degradation over a broad range of usage patterns which eventually results in the catch-all 70% loss after 8y/100k miles warranty.

The truck has been out for years, and I assume all relevant tests will have been done before trucks are sold.

Am I missing anything?

Also, @MickeyAO , if you don't mind me asking, (and this is just my profound ignorance) what (or why) is being tested now that was not tested before by SK/Ford?

As far as why am I testing this cell? We are a completely independent testing house that does testing for a consortium to report actual data. This may end up confirming or contradicting a manufactures results.

 

[MickeyAO]

So based on your data are you saying it is recommended to be in the 20% to 70% range? I only went below 20 once to 15% and only went to 100% once for a trip. Any concerns there or just minimize those as much as possible? Thanks

If you read my previous post on how to make the cells last virtually forever, you will see that I regularly charge to 85% SOC every day. This is just about 80% of actual capacity.

 

[DesertEV]

If you read my previous post on how to make the cells last virtually forever, you will see that I regularly charge to 85% SOC every day. This is just about 80% of actual capacity.

Got it. Go 85% and don’t go below 20%

 

[RyZac]

What's the thought on the Out of Spec Lightning team losing power with truck still showing 12 miles of range?
To me they should do an update where they reduce the overall range by 20 miles and protect the lower portion of the battery in addition to the 10% top range they are protecting.

 

[Box Cat]

We actually do a 95% SOC and a 5% SOC pulse in our HPPC tests just to help calibrate the settings for our cycle life testing. The 0% SOC step is to find the voltage curve to Vmin. A pulse at 0% SOC would drive the cell below Vmin, so you don't want or need that pulse, nor 100% SOC in the HPPC test.

I have static capacity results for different C rates and temperatures, but nothing really stands out as concerning.

You will notice that I specifically state I don't want to go below actual 20% SOC due to what I saw.




As far as why am I testing this cell? We are a completely independent testing house that does testing for a consortium to report actual data. This may end up confirming or contradicting a manufactures results.

I see, thanks. Something like keeping them in check.

By the way, from what I read in the DOE document they don’t drive the cell below Vmin0, they just reach Vmin0 at the end of the test.

Out of curiosity, and if you are at liberty to say, how many cells are you testing?

Sponsored