PicoScope 7 Automotive
Available for Windows, Mac, and Linux, the next evolution of our diagnostic scope software is now available.
EV HV Battery SOC, Capacity & SOH
-
Steve Smith
- Pico Staff Member
- Posts: 1671
- Joined: Sun Aug 25, 2013 7:22 am
EV HV Battery SOC, Capacity & SOH
Following the recent discovery of a rapid access feature to acquire HV battery state of health (SOH) using our generic scan tool, I was alarmed to see our 2016 e-Golf return a value of 52.2%
Understanding how this SOH value is derived would be something of a dark art as I am sure it is hidden in the depths of an algorithm which takes into account usage, charging styles (fast/slow) temperature, load, etc. over many years. Please note, our e-Golf covers very little in the way of mileage and serves purely as a road legal training vehicle with limited range
In addition to the SOH value above, the scan tool also informs “With an increasing number of charging cycles and in increase in calendric aging, the SOH of the HV battery decreases. After approx. 10 years the remaining capacity of the HV battery can still be approx. 70%”
Having queried several EV’s for their HV battery SOH with our scan tool, the statement above varies; for example, recent testing of a 2018 Renault Zoe informed of an expected “HV battery capacity of approximately 81% after 10 years”
The following exercise was carried out in a controlled environment capturing HV battery voltage at the Power and Control Electronics unit (JX1) of our e-Golf using a differential probe (https://www.picoauto.com/products//elec ... e-x10-x100) and a BNC+ 2000 A current clamp (https://www.picoauto.com/products/curre ... rent-clamp) around the HV DC battery cable
With the above probes in place, we capture the HV battery voltage and current flow during charging and discharging
Starting with our e-Golf HV battery SOC at 10% and a range of 4 miles (see below) the HV battery was charged overnight via a 2.2 kW / Mode 2 charger (Fondly referred to as a Granny charger)
Before I go on, be aware of SOC values displayed by various scan tools! On the e-Golf, our generic scan tool returned 16% SOC whilst VCDS returned 10%!
Below we capture the volage and current values during an overnight charge
Note we have also added a math channel “A*B” which is Voltage x Current to derive Power in Watts
Above we can see the total charge time 8 hours and 23 minutes at approx. 2 kW.
Charge time 8 hour 23 minutes (503 minutes)
503 / 60 = 8.383 hours
Mean charge to HV battery based on the above capture:
2033 W x 8.383 Hours = 17,042.639 / 1000 = 17.043 kWh
This data in itself is a good indication of battery condition based on charge time alone. Why?
Let’s do some quick math based on expected charge time:
The e-Golf has a 21.2 kWh HV battery
Expected charge time 0% to 100 % = Battery Capacity / Max charge power
21.2 kWh / 2.033 kW = 10.428 hours for 0% to 100 % SOC
Given our e-Golf HV battery SOC = 10% prior to charging:
21.2 – 10% = 19.08 kWh / 2.033 kW = 9.385-hour “Expected” charge time.
Our e-Golf battery was fully charged in 8h 23 min, we therefore charged approx. 1 hour shorter than expected. Allowing for erroneous values from serial data (SOC) the actual charge time suggests our battery is performing as expected (not charging too soon or too long)
A HV battery with a specified capacity of 21.2 kWh, that transitions from a SOC of 10% to 100% in 2- hours (when charged at 6 A) is an indicator of a failing battery with reduced capacity
Moving on:
Capacity and SOH calculations based on voltage and current measurements
Guided by the information found here https://www.reddit.com/r/electricvehicl ... _and_real/ we use the following abbreviations in order to simply the formulas
1. The Battery charge Percentage when you start charging (SP)
2. The Battery charge percentage when you finish charging (EP)
3. The amount of kWh the charger used in the charging session. (TC)
4. The original battery capacity of the EV as stated by the manufacturer. (OB)
5. Measured capacity calculated from physically charging the HV battery (MC)
The measured capacity of our HV battery (MC) = 100 * TC / (EP-SP)
Amount of kWh the charger used (TC):
2033 W x 8.383 Hours = 17,042.639 / 1000 = 17.043 kWh (Calculated at the HV battery)
100 * 17.043 kWh / (100 %– 10 %)
1704.3 / 90
Measured capacity (MC): = 18.936 kWh
MC expressed as a percentage:
Specified capacity = 21.2 kWh
Actual capacity 18.936 / 21.2 = 89.321 % of specified
Our e-Golf battery capacity has reduced by 10.679 %
SOH
100 * MC/OB
100 * 18.936 / 21.2
89.321 % (Not as stated by generic scan tool at 52.2 % unless other items are factored into the Algorithm)
Please note, this battery is now 8 years old
The above maths can also be applied in reverse where a fully charged HV battery is discharged whilst driving. In the test carried out below, the vehicle started a 54-mile journey with a SOC @ 100 % and completed the journey with a SOC @ 26 %. Note, during this journey, regeneration was only applied during braking and not by selecting regeneration via over-run (I.e., when foot removed from the accelerator pedal)
Before I go on, the numbered circles in the image above denote various driving styles that are worth a mention when we analyse the “trend” of the battery voltage on channel A
1. Motorway driving at 75 mph for approx. 10 miles without cabin heating (Note rapid decline in battery voltage)
2. Motorway driving at 60 mph for approx. 37 miles without cabin heating
3. Visit McDonalds Drive through (for a hot drink)
4. Single lane “A” road driving with light to medium traffic (with cabin heating) for approx. 7 miles
Now to the maths using the “Mean” kW value during our drive time of 1.2 hour
Measured drive time approx. 1 hour 12 minutes (= 72 minutes)
72 / 60 = 1.2 hour
The measured capacity of our HV battery (MC):
100 * Mean battery energy consumed / (SOC prior to road test – SOC after road-test)
Mean kWh consumed during road-test (12.41 kW)
12410 W x 1.12 Hours = 13,899.2 / 1000 = 13.8992 kWh (Calculated at the HV battery)
100 * 13.8992 kWh / (100 % – 26 %)
1,389.92 / 74
Measured capacity (MC): = 18.783 kWh
MC expressed as a percentage:
Specified capacity = 21.2 kWh
Actual capacity 18.783 / 21.2 = 88.599 % of specified
Our e-Golf battery capacity has reduced by 11.4 %
SOH
100 * MC/OB
100 * 18.783 / 21.2
88.599 % (Not as stated by generic scan tool at 52.2 % unless other items are factored into the Algorithm)
Please note, this battery is now 8 years old
As we can see from the above data, whether we are charging or discharging our HV battery, using the mean kW values derived from voltage and current returns very similar values. Measuring direct at the HV battery (where access permits) removes the variables of serial data interpretation between scan tools and losses to other circuits & components when attempting similar measurements based on the RMS mains current during charging!
Note: Not all mains current is delivered to the HV battery during charging, please see the following EV Guided test (GT): Click on Guided tests > Electric vehicles > Charger-vehicle tests > Charging current split > Guide and settings file (see our GT example waveform below)
To conclude
What has become apparent during these tests is just how inaccurate the predicted ranges (mileage) can be when starting your journey.
Our e-Golf claims a range of 106 miles with a 100% SOC, after driving 54 miles our SOC was 26 % which translates to a displayed range of 18 miles (54 + 18 = 72, an error or 34%)
Of course, we must take into account the way the vehicle is driven, the consumers used (i.e. heating) and the amount of regeneration applied to the battery. I know if I modified my driving style and used regeneration level 3 (which applies max regen during over-run) I could have increased the range
I hope this helps, take care……..Steve
- 1
In addition to the SOH value above, the scan tool also informs “With an increasing number of charging cycles and in increase in calendric aging, the SOH of the HV battery decreases. After approx. 10 years the remaining capacity of the HV battery can still be approx. 70%”
Having queried several EV’s for their HV battery SOH with our scan tool, the statement above varies; for example, recent testing of a 2018 Renault Zoe informed of an expected “HV battery capacity of approximately 81% after 10 years”
The following exercise was carried out in a controlled environment capturing HV battery voltage at the Power and Control Electronics unit (JX1) of our e-Golf using a differential probe (https://www.picoauto.com/products//elec ... e-x10-x100) and a BNC+ 2000 A current clamp (https://www.picoauto.com/products/curre ... rent-clamp) around the HV DC battery cable
With the above probes in place, we capture the HV battery voltage and current flow during charging and discharging
Starting with our e-Golf HV battery SOC at 10% and a range of 4 miles (see below) the HV battery was charged overnight via a 2.2 kW / Mode 2 charger (Fondly referred to as a Granny charger)
- 2
Below we capture the volage and current values during an overnight charge
Note we have also added a math channel “A*B” which is Voltage x Current to derive Power in Watts
- 3
Charge time 8 hour 23 minutes (503 minutes)
503 / 60 = 8.383 hours
Mean charge to HV battery based on the above capture:
2033 W x 8.383 Hours = 17,042.639 / 1000 = 17.043 kWh
This data in itself is a good indication of battery condition based on charge time alone. Why?
Let’s do some quick math based on expected charge time:
The e-Golf has a 21.2 kWh HV battery
Expected charge time 0% to 100 % = Battery Capacity / Max charge power
21.2 kWh / 2.033 kW = 10.428 hours for 0% to 100 % SOC
Given our e-Golf HV battery SOC = 10% prior to charging:
21.2 – 10% = 19.08 kWh / 2.033 kW = 9.385-hour “Expected” charge time.
Our e-Golf battery was fully charged in 8h 23 min, we therefore charged approx. 1 hour shorter than expected. Allowing for erroneous values from serial data (SOC) the actual charge time suggests our battery is performing as expected (not charging too soon or too long)
A HV battery with a specified capacity of 21.2 kWh, that transitions from a SOC of 10% to 100% in 2- hours (when charged at 6 A) is an indicator of a failing battery with reduced capacity
Moving on:
Capacity and SOH calculations based on voltage and current measurements
Guided by the information found here https://www.reddit.com/r/electricvehicl ... _and_real/ we use the following abbreviations in order to simply the formulas
1. The Battery charge Percentage when you start charging (SP)
2. The Battery charge percentage when you finish charging (EP)
3. The amount of kWh the charger used in the charging session. (TC)
4. The original battery capacity of the EV as stated by the manufacturer. (OB)
5. Measured capacity calculated from physically charging the HV battery (MC)
The measured capacity of our HV battery (MC) = 100 * TC / (EP-SP)
Amount of kWh the charger used (TC):
2033 W x 8.383 Hours = 17,042.639 / 1000 = 17.043 kWh (Calculated at the HV battery)
100 * 17.043 kWh / (100 %– 10 %)
1704.3 / 90
Measured capacity (MC): = 18.936 kWh
MC expressed as a percentage:
Specified capacity = 21.2 kWh
Actual capacity 18.936 / 21.2 = 89.321 % of specified
Our e-Golf battery capacity has reduced by 10.679 %
SOH
100 * MC/OB
100 * 18.936 / 21.2
89.321 % (Not as stated by generic scan tool at 52.2 % unless other items are factored into the Algorithm)
Please note, this battery is now 8 years old
The above maths can also be applied in reverse where a fully charged HV battery is discharged whilst driving. In the test carried out below, the vehicle started a 54-mile journey with a SOC @ 100 % and completed the journey with a SOC @ 26 %. Note, during this journey, regeneration was only applied during braking and not by selecting regeneration via over-run (I.e., when foot removed from the accelerator pedal)
- 4
1. Motorway driving at 75 mph for approx. 10 miles without cabin heating (Note rapid decline in battery voltage)
2. Motorway driving at 60 mph for approx. 37 miles without cabin heating
3. Visit McDonalds Drive through (for a hot drink)
4. Single lane “A” road driving with light to medium traffic (with cabin heating) for approx. 7 miles
Now to the maths using the “Mean” kW value during our drive time of 1.2 hour
Measured drive time approx. 1 hour 12 minutes (= 72 minutes)
72 / 60 = 1.2 hour
The measured capacity of our HV battery (MC):
100 * Mean battery energy consumed / (SOC prior to road test – SOC after road-test)
Mean kWh consumed during road-test (12.41 kW)
12410 W x 1.12 Hours = 13,899.2 / 1000 = 13.8992 kWh (Calculated at the HV battery)
100 * 13.8992 kWh / (100 % – 26 %)
1,389.92 / 74
Measured capacity (MC): = 18.783 kWh
MC expressed as a percentage:
Specified capacity = 21.2 kWh
Actual capacity 18.783 / 21.2 = 88.599 % of specified
Our e-Golf battery capacity has reduced by 11.4 %
SOH
100 * MC/OB
100 * 18.783 / 21.2
88.599 % (Not as stated by generic scan tool at 52.2 % unless other items are factored into the Algorithm)
Please note, this battery is now 8 years old
As we can see from the above data, whether we are charging or discharging our HV battery, using the mean kW values derived from voltage and current returns very similar values. Measuring direct at the HV battery (where access permits) removes the variables of serial data interpretation between scan tools and losses to other circuits & components when attempting similar measurements based on the RMS mains current during charging!
Note: Not all mains current is delivered to the HV battery during charging, please see the following EV Guided test (GT): Click on Guided tests > Electric vehicles > Charger-vehicle tests > Charging current split > Guide and settings file (see our GT example waveform below)
- EV GT
What has become apparent during these tests is just how inaccurate the predicted ranges (mileage) can be when starting your journey.
Our e-Golf claims a range of 106 miles with a 100% SOC, after driving 54 miles our SOC was 26 % which translates to a displayed range of 18 miles (54 + 18 = 72, an error or 34%)
Of course, we must take into account the way the vehicle is driven, the consumers used (i.e. heating) and the amount of regeneration applied to the battery. I know if I modified my driving style and used regeneration level 3 (which applies max regen during over-run) I could have increased the range
I hope this helps, take care……..Steve
Re: EV HV Battery SOC, Capacity & SOH
Hello,
We are talking about the first generation of e_Golf(May 2014 -June 2016) or the second generation (2016-)?
For the first generation nominal capacity of battery(Lithium-Ion type) is 24.2 kWh(nominal voltage 323v *capacity 75Ah) and give me 76.4% SOH.
For the second generation the nominal capacity of battery is 35.8kWh(useable 32kWh) and give me 52.1%SOH.
I used estimated time for 100% charge,medium charging current and some data from SSP530 for calculation.
Best regards
We are talking about the first generation of e_Golf(May 2014 -June 2016) or the second generation (2016-)?
For the first generation nominal capacity of battery(Lithium-Ion type) is 24.2 kWh(nominal voltage 323v *capacity 75Ah) and give me 76.4% SOH.
For the second generation the nominal capacity of battery is 35.8kWh(useable 32kWh) and give me 52.1%SOH.
I used estimated time for 100% charge,medium charging current and some data from SSP530 for calculation.
Best regards
-
Steve Smith
- Pico Staff Member
- Posts: 1671
- Joined: Sun Aug 25, 2013 7:22 am
Re: EV HV Battery SOC, Capacity & SOH
Hello Victor, thank you for the feedback
That figure you calculated of 52.1% SOH is intriguingly close to the original value quoted by our scan tool of 52.2% suggesting perhaps our scan tool has identified our HV battery incorrectly as a 35.8 kWh!
Just looking back at the scan tool VIN description, it has returned EEPROM-READ-ERROR for the VIN field so it is possible and would explain the worrying SOH value of 52.2%
The information I have on our e-Golf:
VIN: WVWZZZAUZGW908645
Nominal Battery Energy 21.2 kWh (323 V) Acquired from SSP-530
I hope this helps, take care……Steve
That figure you calculated of 52.1% SOH is intriguingly close to the original value quoted by our scan tool of 52.2% suggesting perhaps our scan tool has identified our HV battery incorrectly as a 35.8 kWh!
Just looking back at the scan tool VIN description, it has returned EEPROM-READ-ERROR for the VIN field so it is possible and would explain the worrying SOH value of 52.2%
The information I have on our e-Golf:
VIN: WVWZZZAUZGW908645
Nominal Battery Energy 21.2 kWh (323 V) Acquired from SSP-530
I hope this helps, take care……Steve