How much electricity, really?

[Porsche]

I have not yet, but am planning on getting a wifi charger so that I can get a reduced electrical rate from my local utility. But I am curious. Has anyone out here compared the actual, measured kWh supplied from the utility with the expended kWh driving? I make all sorts of observations about the car's efficiency, but I'm curious to know the losses from charging (including consumption while just sitting). Please, I don't need to know about published charging efficiency, etc. I want to know if anyone has actually measured this, total kWh from the wall vs. total kWh driving.

 

[ZNuts]

Well, what would you rather believe? Some random guy on the internet without much electrical knowledge or test equipment or properly conducted test?

 

[Porsche]

Come on, special equipment? With a wifi enabled charger (which I would imagine nearly everyone has), you can get a detailed report of your car's exact energy consumption at the wall socket. Your power company will even send you one. The car's Energy Details display will give you the total energy since last full charge from the car's perspective. It doesn't take an engineer to figure it out. Also, have you browsed this forum? There are a ton of really helpful, really smart people out here, posting all sorts of detailed analyses.

 

[ZNuts]

I do stuff like that professionally. It's easy to take data and also very easy to miss details and not accounting for things. Very often we get data that just doesn't match expectation or confirm with other past track record and take a lot effort to find out why. Get good repeatable data is not trivial.

Meanwhile there are quite a lot decent published info out there already. What I'm saying is what's wrong with those info vs someone here doing their own "study"?

 

[Infinion]

I have not yet, but am planning on getting a wifi charger so that I can get a reduced electrical rate from my local utility. But I am curious. Has anyone out here compared the actual, measured kWh supplied from the utility with the expended kWh driving? I make all sorts of observations about the car's efficiency, but I'm curious to know the losses from charging (including consumption while just sitting). Please, I don't need to know about published charging efficiency, etc. I want to know if anyone has actually measured this, total kWh from the wall vs. total kWh driving.

Hey Porche, been wanting to make time to get into this post for a while and collect some initial data for the discussion. I have a Kill-a-Watt which is severely miscalibrated, so did the next best thing and grabbed a wifi switch (claimed to be UL listed on Amazon) that measures power stats.

But I am curious. Has anyone out here compared the actual, measured kWh supplied from the utility with the expended kWh driving?
[...] I'm curious to know the losses from charging (including consumption while just sitting). P

The driving part of the comparison would be really tricky since the activity is a variable C rate discharge depending on us and our environment's discretion, but if we're keeping the discussion loosely around power consumption and conversion efficiency not related to the motor, it could prove insightful.

I can start by sharing reported wall plug power vs the car's reported input and output power on both ends of the onboard charger (through TorquePro and the bolt PIDs).

I conducted a roughly 3.36-hour charge and captured beginning and end screenshots of the wifi switch app, and the Torque Pro app to note voltage, current, power, and total kWh stats. These stats were used together to make a final efficiency = Power_out/Power_in (or η = 1 - Ploss/Pin) calculation to find the wall-plug efficiency to charge the battery (at my local temperature). I also took a few thermal images of the EVSE and the wifi switch (relay) and EVSE enclosures and wires after they had been running for a while. A 12 AWG 25ft extension was used just so I could connect the WiFi plug. Slight impact on delivered power but for all intents and purposes, I'm only going to consider losses AFTER the switch and ignore the extension and lines to the utility.

I guess this is a good time to point out that keeping the wires wrapped in a bundle around the EVSE is not such a great thing. Yeah, you save time, but you're impacting the passive cooling of the wires and forcing every adjacent cable to elevate in temperature because there's less cooler air to reject heat into. This was in a carport but Imagine charging the car with the EVSE in the Summer noon sun. I bet it would reach up to 70ºC or 80ºC. Thats also to say that hotter wires have higher resistance (PTC) in other words ... more losses!


So onto my little test, the daily power consumption on my smart switch already has 0.08 kWh from the appliance it normally connects with and should be subtracted.
Somewhere around 21 minutes after charging at 12A, I took the first screenshots. In the WiFi switch app, we see the voltage, current, and power at the wall plug. It's important to note that these readings are before the EVSE, so we can note what the full delivered power at the wall plug is.


In the TorquePro app, we can note my SoC, my battery capacity estimate, reported AC input volts and amps (CAC volts and amps located at lower right side), on-board charger DC-DC input and DC-DC output (DCC lower left). We can also compare the DCC power output with batt volts multiplied by CDC output amps and see by how much they disagree as a result of calculation or measurement error.



At this moment 0.4 kWh into charging,

• This is a 0.0697 C or 1/15 C charge
• Wall plug power measured 1336.4 Watts, 107.4 V, 12.392 A
• CAC volts and amps likely means we're reading the volts and amps before the AC-DC full bridge rectifier and filtering stage
• CAC power was reported to be 1248.0 Watts, 104 V, 12 A. There is slight disagreement between current, and this could be a measurement error or current diverted to the control and power electronics not used in charging.
• Reasonable to say there is an 88.4 W loss (system efficiency now 93.3 %) in power between the wall plug and the filtering stage of the on-board charger.
• DC-DC input power was reported to be 1099 W. This means there is an 149 W loss in power in the rectification and power factor correction stage (88 % efficient AC-DC conversion, system efficiency now 82.2 %)
• DC-DC output power reported to be 1058 W. DC-DC efficiency is 96.2 %, and system efficiency is now 79.2 %
• Using Batt Volts and CDC amps instead of DCC PowerOut PIDs, output power is instead 1079.166 W, implying DC-DC efficiency of 97.9 % and system efficiency of 80.5 %. You'd think this power would be lower than the DC-DC output, but that's not what's reported by the Bolt PIDs. Could be a measurement or calculation error, I don't really have a great explanation for this.

At the end of the 3 hour test,

• Wall Plug power reduced to 1273.5 W, 108.7 V, 12.207 A
• CAC stats, 1206.4 W, 104 V, 11.6 A. Implies cable and EVSE losses of 67.1 W, section and system efficiency of 94.7 % from here
• DC-DC input power 1090 W, loss of 116.4 W, section efficiency of 90.4 %, system efficiency 85.6 %
• DC-DC output power (DCC PID) 1052 W, loss of 38 W, section efficiency of 96.5 %, system efficiency of 82.6 %
• DC-DC output power (Batt Volts * HV Amps) 1016.65 W. Loss of 73.35 W, section efficiency 96.5 %, system efficiency 79.8 %. However, we see a difference of 0.05 A between CDC and HV amps, meaning 0.05 A*356.72 V = 17.8 W could be going to the coolant pumps if we compare the power difference of the DC-DC output ,and the rest of the loss is from the DC-DC converter

In the last set of screenshots, I compare the WiFi switch kWh to the PIDs' two kWh readings.

• Wall plug kWh = 4.58 - 0.08 = 4.5 kWh
• Last Chrg Wh shows 4.01 kWh, which seems like it could be counted at the input to the on-board charger. I say that because comparing the WiFi switch kWh to this reading gives a section efficiency of 89 % which is pretty close but worse than what I got while reading the DCC power input PIDs during the charge.
• Chrg HV Last shows 3.15 kWh as the energy actually getting to the batteries. This implies wall plug to batteries system efficiency of 70 % which is... pretty horrible and a very big departure from the power readings during the test.

So the most pessimistic efficiency figure would be comparing kWh of the whole charging session to get efficiency. I sort of lean towards that more pessimistic efficiency because it's probably coming from the "gas gauge" coulomb counter of the BMS and would naturally take into consideration power being taken from the pumps, fans, car battery charging, contactors, and the rest of the 12V systems. However, I still want to run another test covering a larger charge to get better numbers. I also wouldn't mind doing a test in the winter to see how bad this really gets when the battery heater needs to be used.

I remember some time in January that I jotted down some estimates for power consumption during the Winter in January.
I probably used the Inst P (KW), Batt Volts, and CDC Amps vs HV Amps PIDs to estimate idle power consumption while charging.
It says:

• Battery Pump: 39 Watts
• Battery Heater Core 2445W
• Cabin fan 100% 113.57W using CDC or 136W using HV amps 2
• Rear Def 200W

Also when fully charged, the contactor for the EVSE disconnects power, and to measure tiny power consumption using an OBD dongle is tricky because amperage resolution on the batteries is 50mA. In other words you can't see anything smaller than about 20W.

 

[LostinEVland]

I have not yet, but am planning on getting a wifi charger so that I can get a reduced electrical rate from my local utility. But I am curious. Has anyone out here compared the actual, measured kWh supplied from the utility with the expended kWh driving? I make all sorts of observations about the car's efficiency, but I'm curious to know the losses from charging (including consumption while just sitting). Please, I don't need to know about published charging efficiency, etc. I want to know if anyone has actually measured this, total kWh from the wall vs. total kWh driving.

Hey Porche, been wanting to make time to get into this post for a while and collect some initial data for the discussion. I have a Kill-a-Watt which is severely miscalibrated, so did the next best thing and grabbed a wifi switch (claimed to be UL listed on Amazon) that measures power stats.

But I am curious. Has anyone out here compared the actual, measured kWh supplied from the utility with the expended kWh driving?
[...] I'm curious to know the losses from charging (including consumption while just sitting). P

The driving part of the comparison would be really tricky since the activity is a variable C rate discharge depending on us and our environment's discretion, but if we're keeping the discussion loosely around power consumption and conversion efficiency not related to the motor, it could prove insightful.

I can start by sharing reported wall plug power vs the car's reported input and output power on both ends of the onboard charger (through TorquePro and the bolt PIDs).

I conducted a roughly 3.36-hour charge and captured beginning and end screenshots of the wifi switch app, and the Torque Pro app to note voltage, current, power, and total kWh stats. These stats were used together to make a final efficiency = Power_out/Power_in (or η = 1 - Ploss/Pin) calculation to find the wall-plug efficiency to charge the battery (at my local temperature). I also took a few thermal images of the EVSE and the wifi switch (relay) and EVSE enclosures and wires after they had been running for a while. A 12 AWG 25ft extension was used just so I could connect the WiFi plug. Slight impact on delivered power but for all intents and purposes, I'm only going to consider losses AFTER the switch and ignore the extension and lines to the utility.

I guess this is a good time to point out that keeping the wires wrapped in a bundle around the EVSE is not such a great thing. Yeah, you save time, but you're impacting the passive cooling of the wires and forcing every adjacent cable to elevate in temperature because there's less cooler air to reject heat into. This was in a carport but Imagine charging the car with the EVSE in the Summer noon sun. I bet it would reach up to 70ºC or 80ºC. Thats also to say that hotter wires have higher resistance (PTC) in other words ... more losses!


So onto my little test, the daily power consumption on my smart switch already has 0.08 kWh from the appliance it normally connects with and should be subtracted.
Somewhere around 21 minutes after charging at 12A, I took the first screenshots. In the WiFi switch app, we see the voltage, current, and power at the wall plug. It's important to note that these readings are before the EVSE, so we can note what the full delivered power at the wall plug is.


In the TorquePro app, we can note my SoC, my battery capacity estimate, reported AC input volts and amps (CAC volts and amps located at lower right side), on-board charger DC-DC input and DC-DC output (DCC lower left). We can also compare the DCC power output with batt volts multiplied by CDC output amps and see by how much they disagree as a result of calculation or measurement error.



At this moment 0.4 kWh into charging,

• This is a 0.0697 C or 1/15 C charge
• Wall plug power measured 1336.4 Watts, 107.4 V, 12.392 A
• CAC volts and amps likely means we're reading the volts and amps before the AC-DC full bridge rectifier and filtering stage
• CAC power was reported to be 1248.0 Watts, 104 V, 12 A. There is slight disagreement between current, and this could be a measurement error or current diverted to the control and power electronics not used in charging.
• Reasonable to say there is an 88.4 W loss (system efficiency now 93.3 %) in power between the wall plug and the filtering stage of the on-board charger.
• DC-DC input power was reported to be 1099 W. This means there is an 149 W loss in power in the rectification and power factor correction stage (88 % efficient AC-DC conversion, system efficiency now 82.2 %)
• DC-DC output power reported to be 1058 W. DC-DC efficiency is 96.2 %, and system efficiency is now 79.2 %
• Using Batt Volts and CDC amps instead of DCC PowerOut PIDs, output power is instead 1079.166 W, implying DC-DC efficiency of 97.9 % and system efficiency of 80.5 %. You'd think this power would be lower than the DC-DC output, but that's not what's reported by the Bolt PIDs. Could be a measurement or calculation error, I don't really have a great explanation for this.

At the end of the 3 hour test,

• Wall Plug power reduced to 1273.5 W, 108.7 V, 12.207 A
• CAC stats, 1206.4 W, 104 V, 11.6 A. Implies cable and EVSE losses of 67.1 W, section and system efficiency of 94.7 % from here
• DC-DC input power 1090 W, loss of 116.4 W, section efficiency of 90.4 %, system efficiency 85.6 %
• DC-DC output power (DCC PID) 1052 W, loss of 38 W, section efficiency of 96.5 %, system efficiency of 82.6 %
• DC-DC output power (Batt Volts * HV Amps) 1016.65 W. Loss of 73.35 W, section efficiency 96.5 %, system efficiency 79.8 %. However, we see a difference of 0.05 A between CDC and HV amps, meaning 0.05 A*356.72 V = 17.8 W could be going to the coolant pumps if we compare the power difference of the DC-DC output ,and the rest of the loss is from the DC-DC converter

In the last set of screenshots, I compare the WiFi switch kWh to the PIDs' two kWh readings.

• Wall plug kWh = 4.58 - 0.08 = 4.5 kWh
• Last Chrg Wh shows 4.01 kWh, which seems like it could be counted at the input to the on-board charger. I say that because comparing the WiFi switch kWh to this reading gives a section efficiency of 89 % which is pretty close but worse than what I got while reading the DCC power input PIDs during the charge.
• Chrg HV Last shows 3.15 kWh as the energy actually getting to the batteries. This implies wall plug to batteries system efficiency of 70 % which is... pretty horrible and a very big departure from the power readings during the test.

So the most pessimistic efficiency figure would be comparing kWh of the whole charging session to get efficiency. I sort of lean towards that more pessimistic efficiency because it's probably coming from the "gas gauge" coulomb counter of the BMS and would naturally take into consideration power being taken from the pumps, fans, car battery charging, contactors, and the rest of the 12V systems. However, I still want to run another test covering a larger charge to get better numbers. I also wouldn't mind doing a test in the winter to see how bad this really gets when the battery heater needs to be used.

I remember some time in January that I jotted down some estimates for power consumption during the Winter in January.
I probably used the Inst P (KW), Batt Volts, and CDC Amps vs HV Amps PIDs to estimate idle power consumption while charging.
It says:

• Battery Pump: 39 Watts
• Battery Heater Core 2445W
• Cabin fan 100% 113.57W using CDC or 136W using HV amps 2
• Rear Def 200W

Also when fully charged, the contactor for the EVSE disconnects power, and to measure tiny power consumption using an OBD dongle is tricky because amperage resolution on the batteries is 50mA. In other words you can't see anything smaller than about 20W.

Holy hell some of you guys in here make me feel stupid thanks for posting this

 

[Infinion]

Holy hell some of you guys in here make me feel stupid thanks for posting this

Not at all man, all the data is definitely a lot to take in if you don't know what you're looking for. I'm a numbers junkie, and I do these kinds of calculations regularly for EE and circuit board design work, so I don't get too much of a headache looking at all this, not to mention it's kind of fun.

I just wanted to update that I recorded a 10%-98% charge test. Unfortunately, I didn't get the Wh report from the car on how much went into the battery pack, but hopefully using the battery capacity estimate is comparably accurate.

Torque Pro screenshots of beginning and end charge from 10% to 98%:

Smart switch energy monitor:

Smart switch reported 17.51 kWh were passed at the wall plug
My battery's most recent capacity estimate is 14.56 kWh as of yesterday after this charge (not shown), so I'll use this number since it was updated immediately after the charge.
The session charged 98.001% - 10.842% = 87.159% as seen in TorquePro.
Therefore using the battery capacity, 14.56 kWh * 0.87159 = 12.69 kWh delivered to the battery pack.
This means 17.51 kWh (total) - 12.69 kWh (batteries) = 4.82 kWh (losses). Losses include heat lost from conversion and cables, as well as running the battery + electronics coolant pumps and radiator fan
So the conversion efficiency of a 12A level 1 nearly full charge can be estimated at 12.69 kWh (batteries) / 17.51 (total) * 100% = 72.47%.
That falls pretty closely to the estimate I gave in the last test using the energy monitor and "Chrg HV" PID to find the efficiency last time.

So it's looking more certain that summertime efficiency when batteries are between 21ºC - 28ºC (69.8ºF - 82.4ºF), factoring in the power needed to gently run the active cooling systems, is around 71.2%.

If anyone has a guess as to why the efficiency calcs from the last test are so optimistic, reply with your thoughts. My guess is I'm just taking snapshots when the active cooling is not drawing much power, so the delivered power is a bit higher.

Just to be completionist, I'll plan to do an 8A test. If anyone wants to reproduce my numbers in their own test, please feel free. This test is really only applicable in my environmental conditions where there is no Sun beating down on the car. It's a pretty ideal summertime efficiency test.

 

[Porsche]

Infinion, thank you for taking up the gauntlet. As usual, an incredible, detailed analysis. I must say though, I was thinking of a much simpler but effective solution.

All you really need is a wi-fi charger (or one of those "kill-a-watt" outlet monitors). I don't have one yet, but my local utility offers big rebates on select wi-fi chargers. Plus, they monitor your exact electrical usage for the car, and bill you at a special rate. So, you know the exact electrical consumption directly from the outlet, to the nearest kWh, on a minute by minute basis. You can see usage when charging, when sitting idle, maintaining battery temperature, etc. Then, all you need to do is record the car's usage from the Energy Details display. You'll know the exact energy in from the outlet, and the energy used by the car for the same period of time. Start with a full charge. drive the car. record kWh used. Charge again. Check energy from wall for the charge. Calculate loss due to charging. Concerned that you might not have charged up to the exact same state of charge? simply repeat 10 or 20 times. Plus, you can, separately (or together), keep track of the parasitic energy supplied when the car's just sitting there. If you want, you can also get mileage from the odometer to see your true eMPG from the wall.