Li-ion battery health measurements by reading the Battery Pack Capacity signal

[JustAnotherPoorDriver]

I think the consensus is the the battery pack is the Blue Energy EHW5 or EHW5B model used in the Accord Hybrid and the Clarity Fuel Cell. See this Accord Hybrid teardown and the product info page with a list of Honda vehicles which use the pack.

Apparently the Accord is Blue Energy, the Clarity is Panasonic if this is unchanged since they announced it in 2017

See the top of this page, which translates a Japanese article
https://tech.nikkeibp.co.jp/dm/atcl...KWjSVqRjgffC5EPRvhOhp8m6-dh2eDfbpaYa4Rr9R54Jo

 

[HagerHedgie]

Does anyone think it’s worth it to stop charging at 95% instead of 100% (per the Honda link app) to improve battery life?
I would like to keep the car for 300,000miles + if I can.
I’ve heard Honda recommends charging fully for balancing so I’m wondering if this strategy could backfire. Maybe do a 100% every few charges?.?
This seems to avoid the annoying auto start regen feature also

 

[JustAnotherPoorDriver]

Does anyone think it’s worth it to stop charging at 95% instead of 100% (per the Honda link app) to improve battery life?
I would like to keep the car for 300,000miles + if I can.
I’ve heard Honda recommends charging fully for balancing so I’m wondering if this strategy could backfire. Maybe do a 100% every few charges?.?
This seems to avoid the annoying auto start regen feature also

Charging to 100% isn't charging to 100%--Honda has a buffer at both the top and bottom for battery longevity. How much of a buffer exactly (and the properties of the Panasonic batteries) will effect this, so I doubt anyone but Honda can answer your questions with more than a guess...

 

[jdonalds]

On Oct 2 we traded in our 2018 Crimson Clarity Touring PHEV for a new 2020 White version. The new car measured 55Ah battery capacity.

We bought the Crimson car Dec 5 2017 but didn't get a battery capacity measurement then. For the next months battery behavior was as expected, then early in 2019 (about April) the EV Range dropped from the high in the 46-52 range down to 38-40 and stayed there until we traded the car last Friday. Honda wasn't going to do anything and the battery was within spec. It really bugged me that the car had this problem and there was nothing I could do about it. With the range dropping in the winter there were days when the ICE had to kick on which I also hated. Our driving pattern did not change through all of this. The Crimson had about 53K miles at the time of the trade. We averaged about 18K miles per year - quite a bit less due to COVID-19. First two years were more like 22K miles/year.

The new car is showing EV Range of 45-46 and I expect it to climb a bit more.

 

[Ss84]

I just got a report on my car today: 46.8Ah on a 2018 PHEV bought in December 2017 (so not quite 3 years old). It has 28,100 miles and it feels like I've lost a lot of range. I used to be able to get to work and back (just over 44 miles) easily in summer, and now it takes effort to barely make it. I've contacted Honda about it but I'm not really expecting anything to come out of it : /. Seems like an unacceptably high loss if it started at 55Ah (which I'm not sure about). I only asked for the report today because it "felt" like the range was getting noticeably bad and the report seems to confirm it.

 

[jdonalds]

I just got a report on my car today: 46.8Ah on a 2018 PHEV bought in December 2017 (so not quite 3 years old). It has 28,100 miles and it feels like I've lost a lot of range. I used to be able to get to work and back (just over 44 miles) easily in summer, and now it takes effort to barely make it. I've contacted Honda about it but I'm not really expecting anything to come out of it : /. Seems like an unacceptably high loss if it started at 55Ah (which I'm not sure about). I only asked for the report today because it "felt" like the range was getting noticeably bad and the report seems to confirm it.

I had the same problem but my range fell close to 10 miles less overnight and never returned. I expected degradation but not so front loaded. It was still in spec so Honda wouldn't even acknowledge that it was rapid. That bothered me quite a bit. I like the car so much we traded it in, switched colors, and now have the better electric range. I know it is likely to drop over the next two or three years like the first one (purchased Dec 5, 2017, 52K miles) but I'm good with it for a while. I'm hoping a decently priced BEV will hit the market that has less than 20 minutes charge 20-80% To me it's all about charge time for long trips, not range. The other hope is the battery capacity will fall low enough that Honda has to replace it. With a fresh battery I'd keep the car another three or more years.

 

[DucRider]

I'm hoping a decently priced BEV will hit the market that has less than 20 minutes charge 20-80% To me it's all about charge time for long trips, not range.

The 2014 Spark EV did pretty close to that (and you can get them now for ~$8K), but you do need it in combination with range for long trips. Stopping every 40-50 miles to get an 80% charge is not viable.
It's probably better if you look at charge time per 150 miles (or some other distance), as 80% of the 200 kWh pack in the Hummer EV is very different than 80% of the ~ 80 kWh LR Model 3 (but the ranges for an 80% charge are surprisingly close)

 

[jdonalds]

The 2014 Spark EV did pretty close to that (and you can get them now for ~$8K), but you do need it in combination with range for long trips. Stopping every 40-50 miles to get an 80% charge is not viable.
It's probably better if you look at charge time per 150 miles (or some other distance), as 80% of the 200 kWh pack in the Hummer EV is very different than 80% of the ~ 80 kWh LR Model 3 (but the ranges for an 80% charge are surprisingly close)

I was thinking 250 miles of range, 15-20 minutes charge time 20-80%, and a car that costs less than $35K. I can wait.

 

[jpkik96]

I just got a report on my car today: 46.8Ah on a 2018 PHEV bought in December 2017 (so not quite 3 years old). It has 28,100 miles and it feels like I've lost a lot of range. I used to be able to get to work and back (just over 44 miles) easily in summer, and now it takes effort to barely make it. I've contacted Honda about it but I'm not really expecting anything to come out of it : /. Seems like an unacceptably high loss if it started at 55Ah (which I'm not sure about). I only asked for the report today because it "felt" like the range was getting noticeably bad and the report seems to confirm it.

______________

All - I leased a 2018 Clarity Touring in June 2018 and am deciding whether to buy it off the lease and add a Honda-care warranty before the factory warranty expires or just return to Honda. The purchase price today after tax but before the Honda Care would be approx $19.5M. The Clarity now has 35M miles on it and I asked my local dealer to test the battery to help me with my purchase decision. They just advised the battery measures 48.7 Ah, which assuming it was 55ah new would be approx 11.5% battery degradation in 30 months of driving.

For reference, I am NJ and live at the summit of 500 ft mountain range which I need to drive up every time I drive home. I typically drive easy in "Sport" mode so I can lock on max re-generation. I also charge every night to 100% using a ChargePoint Level 2 charger. Doing so still yields 48-50 miles on the guess-o-meter, and the car has been surprisingly reliable with just routine dealer oil changes. I also changed the tires @ 28M miles to Michelin Cross-Climates.

Appreciate your thoughts if the battery degradation seems normal. Also appreciate your thoughts on the cars I am cross-shopping below due primarily to the lack of a Sunroof and modern infotainment in the Clarity: (2020) or certified 2018-29 Accord Touring or EXL Hybrid; 2.) (2020) or certified 2018-2019 Acura TLX 2.4 Tech; 3.) New Tesla M3 SR+ (In NJ, there is no sales tax on EV's and they are now eligible for a $5M NJ refund); and if Toyota starts delivering in 2021, a RAV 4 Prime XSE (eligible for 7.5M Federal credit and partial NJ refund).

Thank you in advance for your review and reply!

 

[insightman]

Evidently your GoM doesn't know about your Battery Capacity Signal because 48-50 miles of GoM range is great this time of year--especially with the non-OEM tires. Do you like driving your Clarity? That's what would make the decision for me. If you were able to get one of the few RAV4 Primes, you should re-sell it for a big profit and use the money to buy your Clarity.

 

[jpkik96]

Evidently your GoM doesn't know about your Battery Capacity Signal because 48-50 miles of GoM range is great this time of year--especially with the non-OEM tires. Do you like driving your Clarity? That's what would make the decision for me. If you were able to get one of the few RAV4 Primes, you should re-sell it for a big profit and use the money to buy your Clarity.

Insightman - I have enjoyed reading your posts over the last few years so thank you for your review and reply. With respect to the GoM, I definitely drive most times with battery efficiency in mind, trying to minimize the use of the heater during the fall. I really do like driving the Clarity (my other car is a Mustang GT 6 speed Conv.) so in many respects I find the Clarity to be a performance car with respect to efficiency! However, as noted the lack of a Sunroof or upgraded Stereo, and the fact that very few NEW Clarity's have been sold in NJ since last Summer, makes me increasingly concerned about future dealer support. Does 11.5% battery degradation for a 30-month-old Clarity seem significant? If yes, that makes my decision to let the lease expire and just potentially pay for the excess mileage that much easier. Thanks again!

 

[insightman]

Insightman - I have enjoyed reading your posts over the last few years so thank you for your review and reply. With respect to the GoM, I definitely drive most times with battery efficiency in mind, trying to minimize the use of the heater during the fall. I really do like driving the Clarity (my other car is a Mustang GT 6 speed Conv.) so in many respects I find the Clarity to be a performance car with respect to efficiency! However, as noted the lack of a Sunroof or upgraded Stereo, and the fact that very few NEW Clarity's have been sold in NJ since last Summer, makes me increasingly concerned about future dealer support. Does 11.5% battery degradation for a 30-month-old Clarity seem significant? If yes, that makes my decision to let the lease expire and just potentially pay for the excess mileage that much easier. Thanks again!

I agree 11.5% is significant, as is the $19.5K buy-out. Both numbers are right on the edge of what would help me decide whether to buy your Clarity. I hope others chime in with their opinions.

I'll admit I haven't paid close attention to the battery degradation numbers people have reported. I'm waiting for the Autel software update to let me start tracking my Battery Pack Capacity. The $19.5K price is not that much less than some people in CA have been paying for a new Clarity PHEV, so I imagine you could buy a used one for less.

The quality of dealer support varies with each dealer. My dealer has been able to keep my Honda hybrids in good shape for 21 years, so I'm confident they will continue to do so. However, many people have not been happy with the way their dealers have serviced their Clarity PHEVs. Assuming your dealer properly administered the Battery Capacity test, there isn't anything they can do about an 11.5% degradation because Honda sets the replacement threshold at 36.6 Ah (~33.5% degradation) in 8 or 10 years (I don't know the warranty for the battery in NJ).

Regarding your prospective replacements for your Clarity, the Accord has more features, including the sunroof and upgraded audio system you desire, but will you miss driving mostly on EV? The Acura would be much quicker (but you have the Mustang for that), has even more features and is more luxurious, but you'd be out of the EV world completely. I know almost nothing about the Tesla Model 3's sound system but I don't remember seeing one with a true sunroof. Would your Mustang lose its attraction when it became the second fastest car in your garage?

In my mind there's no car that could replace our Clarity. Only the much heavier Polestar 1 can do better, but it costs $160K and "better" is just 52 miles of EV range. I find the Clarity's sound system way better than what I've had before and a sunroof isn't a high priority for me. I look forward to the post-vaccine days when we can use it to go places again.

 

[MrFixit]

Does 11.5% battery degradation for a 30-month-old Clarity seem significant?

Just a point of reference...

My 2018 (purchased November 2018) was initially measured at the dealer to be 55.1 Ah.
After 24 months and 19K miles (~14K are EV miles) the battery capacity is 49.6 Ah. This is about a 10% reduction.
My observation is that the reduction has leveled off and is not continuing to drop as quickly as it did initially.

As such, based on my very limited experience, your 11.5% reduction seems 'nominal' to me.

 

[Clarity_Newbie]

@jpkik96

Have to agree with Mr Fixit...the battery degradation seems nominal at this point.

On a side note...how do the Michelin Cross-Climates drive? Any noticeable noise or handling differences?

Good luck with your choice and keep us posted.

 

[jpkik96]

MrFixit and Clarity_Newbie - thank you both for your further comment on my battery degradation inquiry. With respect to the Cross-Climates, I am extremely happy with them. Noise may be slightly increased but the handling in all weather conditions has dramatically improved over the Michelin Energy Savers! Now if the Clarity just had a sunroof and the infotainment from the Accord EXL and Touring this would make my decision a lot easier....

 

[JCA]

It would be interesting to know how the full battery pack capacity correlates directly to actual charge and range, i.e. how much is "hidden" in the buffers left at the top and bottom. The GOM isn't a great way to measure this, but someone who's had an EVSE that tracks usage *and* routinely charges from 0.0 EV (even better, with an OBDII reader to get the actual SOC) might have data. For example, if when new at 55Ah, a full charge takes a repeatable 14kWh, when degraded to 50Ah does the car only take a repeatable 12.7kWh?

Still a lot of variables to try to isolate -- ambient temperature causing battery cooling systems to run while charging or the fact that 0.0 EV isn't an exact measure (it will use some more of the 10% SOC as needed below that, and won't show any EV range until 2.0 or ~15% SOC). But would be interesting data.

 

[MrFixit]

It would be interesting to know how the full battery pack capacity correlates directly to actual charge and range, i.e. how much is "hidden" in the buffers left at the top and bottom.

To me, this has been adequately demonstrated...
Not with the GOM (as you say, it has a lot of variables), but with the following methodology...

First, as you say, a "full charge" (when new with the dealer-measured capacity of 55.1 Ah), in my case was around 14.5 kWh. Note, this is charging from zero EV miles remaining (using a Level 2 EVSE). Since the Clarity battery is rated at 17 kWh, this demonstrates the 'buffering' is 2.5 kWh (or, around 15%). I think this is roughly 10% at the bottom and 5% at the top.

A cross-check on the total capacity can be done using the 55 Ah figure... If the HV battery voltage is ~310V (average value, across a discharge cycle) - You can simply multiply the 55 Ah * 310V and you get 17.05 kWh.

Finally, as another cross-check, my capacity has gone from 55.1 to 49.6. Meanwhile, my "full charge" as measured by the EVSE has gone down from ~14.5 kWh to ~13.5 kWh. The capacity has dropped ~10% and my measured full charge has dropped more like 7%. This is close enough for me to declare consistency, but one could argue the few percent discrepancy could somehow involve the vehicle tinkering with the buffers?... I dunno.

The only thing that really matters (from a warranty perspective) is the capacity as measured by the vehicle. I (and others) will be keeping an eye on this as time goes on. I will also continue to monitor how much input energy corresponds to a 'full' charge as reported by the EVSE.

 

[DucRider]

First, as you say, a "full charge" (when new with the dealer-measured capacity of 55.1 Ah), in my case was around 14.5 kWh. Note, this is charging from zero EV miles remaining (using a Level 2 EVSE). Since the Clarity battery is rated at 17 kWh, this demonstrates the 'buffering' is 2.5 kWh (or, around 15%). I think this is roughly 10% at the bottom and 5% at the top.

Charging losses are 10% (+/-~1%)
That 14.5 kWh from the plug gives you about 13 kWh in the battery.

 

[MrFixit]

Charging losses are 10% (+/-~1%)
That 14.5 kWh from the plug gives you about 13 kWh in the battery.

OK, then... That implies the 'buffer' is more like 25% (maybe 10% at the bottom and 15% at the top)?
I agree there are charging losses. The confusion factor to me is exactly how these losses are booked.
Is the specified 17 kWh rated capacity just the ideal battery itself, or is it the entire 'system' with the plug (and any charging losses) as the reference point?
I suppose it seems more likely that there is actually a 17 kWh battery, and my initial full charge is only filling it to 13 kWh (the other 4 kWh being the 'buffer').

However, I am pretty sure that the EPA rating (kWh per 100 miles) treats the vehicle as a black box and measures the energy input to the plug. This means that charging losses are embedded within the EPA number...

My basic point remains...
The vehicle itself has measured a 10% drop in battery capacity. I have observed a corresponding 7% drop in the 'full charge' input energy. These two are plausibly close, but you may speculate as to what happened with the missing 3%. The vehicle could have altered it's buffers, or being within 3% could just be measurement tolerance somewhere in these numbers. How the buffers get used (or not) as the battery degrades over time is a Honda mystery. Maybe we can figure it out over time.

Do you have any theories about this 3% discrepancy? I am fairly happy to have a result that is only 3% off (3% of a ~50 mile range is only 1.5 miles after all).

 

[DucRider]

OK, then... That implies the 'buffer' is more like 25% (maybe 10% at the bottom and 15% at the top)?
I agree there are charging losses. The confusion factor to me is exactly how these losses are booked.
Is the specified 17 kWh rated capacity just the ideal battery itself, or is it the entire 'system' with the plug (and any charging losses) as the reference point?
I suppose it seems more likely that there is actually a 17 kWh battery, and my initial full charge is only filling it to 13 kWh (the other 4 kWh being the 'buffer').

However, I am pretty sure that the EPA rating (kWh per 100 miles) treats the vehicle as a black box and measures the energy input to the plug. This means that charging losses are embedded within the EPA number...

My basic point remains...
The vehicle itself has measured a 10% drop in battery capacity. I have observed a corresponding 7% drop in the 'full charge' input energy. These two are plausibly close, but you may speculate as to what happened with the missing 3%. The vehicle could have altered it's buffers, or being within 3% could just be measurement tolerance somewhere in these numbers. How the buffers get used (or not) as the battery degrades over time is a Honda mystery. Maybe we can figure it out over time.

Do you have any theories about this 3% discrepancy? I am fairly happy to have a result that is only 3% off (3% of a ~50 mile range is only 1.5 miles after all).

The specified battery size is calculated by nominal voltage x Ah. Battery capacity is not a constant, but a variable that will go up or down under different conditions (degradation is a permanent decrease and a separate issue). The major factors that influence change are temperature and discharge rate.
Here is an example of how a battery rating is derived (this is for an LG 21700 cell, but the same type of rating/testing is done on al Li Ion batteries):

4.1 Standard Test Condition
4.1.1 Standard Charge
Unless otherwise specified, “Standard Charge” shall consist of charging at constant current of 0.3C.
The cell shall then be charged at constant voltage of 4.20V while tapering the charge current.
Charging shall be terminated when the charging current has tapered to 50mA. For test purposes,
charging shall be performed at 25ºC ± 2ºC.
4.1.2 Standard Discharge
“Standard Discharge” shall consist of discharging at a constant current of 0.2C to 2.50V. Discharging
is to be performed at 25 ºC ± 2 ºC unless otherwise noted (such as capacity versus temperature).​

Both the amount of energy available from a battery and the amount of energy to fully charge it vary with temperature. +/- 3% in non laboratory conditions would not be unusual.
As an example, the above 21700 specs:

-10℃ ≥70% of Wh
0℃ ≥80% of Wh
25℃ ≥100% of Wh
55℃ ≥95% of Wh​

The method that Honda uses to assign the Ah number is different than a discharge test and likely embedded in the BMS. It's accuracy will be different than a charge/discharge test under lab conditions, but will still give a good "over time" picture of battery health. I don't know the specific method that Honda is using, but that could also account for the theoretical number generated by the BMS (or other test equipment) and the "real world" number you are getting.

Battery buffer is a bit harder to define, but it is extremely unlikely they are playing with those. The battery capacity test shows the charge voltage on each cell. IIRC, it appears the target voltage is 4.1, and it is a fair assumption that the cells rated max voltage is 4.2. SOC% should be measured in available Ah, so that would be derived as the area under the curve as the voltage drops at a set discharge rate over time. There is a fairly quick voltage drop under load (the higher the load, the more severe), so defining the upper buffer can be difficult, but a 90% SOC will correspond to a cell voltage of around 4.05 (varies some by the actual cell characteristics)

If you look at the pink line (the specified .2C discharge rate for this cell), and look at the where the plot crosses 10% of the total Ah (~500) you can see the corresponding voltage.

The bottom buffer is harder to define as it is a bit "squishy" in the Clarity PHEV. The car fires the ICE even when there is still some charge available, and under certain conditions it will dip further into the battery.