onboard charger efficiency ?

I think you use a way low efficiency. I monitored energy usage for some time with KillAWatt, while charging at home and (assuming that battery pack capacity is 17kWh) came with the efficiency around 96%

 

I have not seen anything regarding the Clarity, but generally, I have heard estimates of between 78 - 85%

 

The battery pack is 17kwh, but it is not allowed to discharge/charge completely. I have heard figures here of a full charge being between 14 and 14.7. Not sure whether that is what you used in your calculations.

 

I am not saying that is not true, but all that loss would have to dissipate in the form of heat, and it would be a lot of heat which I did not notice.
IF you charge the battery with, let's say 10kWh, it means that around 2kWh would have to be lost in the form of heat. That is like running 1kW heater for 2 hours, and if you did it in the garage I am sure you would notice it.

 

If your batter pack capacity is batCap (17kWh in our case), and you charge from the initial SoC (startSOC) to the final SoC (endSOC) your battery is charged with energy:
batCharge = batCap * (endSOC - startSOC)

If you measure AC (alternate current) energy energyAC with a KillAWatt (or similar device) it should show higher value than batCharge, due to losses.

Efficiency of the charger (AC-DC conversion) can be calculated as:

efficiency = batCharge/energyAC = batCap * (endSOC - startSOC) / energyAC

This could change with a temperature and other conditions, but if you do it enough times, average value should tell you approximate efficiency of the charger, and I am sure it would be much higher that 78-85%

 

I have some data to add to the discussion. I charge through the OEM charger on a 120v circuit, through a Kill-a-Watt and a 50-ft outdoor extension cable (I know, I know, shush).

[The extension cable should incur 0.2008Ω of resistance, for 2.410W@12A. Considering that the charger runs at 1440W (ish), this is a 0.167% efficiency loss, and should not impact the overall efficiency that much.]

Based on the average of 13 charge cycles, a charge from 10% SoC (0 EV miles) to 100% SoC should take 16.9 kWh using my current setup. If the total usable battery capacity of the Clarity is 14.2±0.5 kWh, then that would imply the efficiency of my charging setup is 84.0±3.0% efficient.

There are a few things to note. First, I don't know what the actual usable battery capacity is. I have heard rumors, but no specifications or definite answers. Second, the charging efficiency seems to be slightly better when charging from low SoC. Third, I am unsure what the effects of temperature are.

I am anxious to see how warmer weather affects things. My current wall-to-wheel efficiency is around 2.5 mi/kWh (≈85mpge), which is pretty far from EPA 110mpge. I might miss my CO2 goals if efficiency doesn't pick up soon.

 

Not sure if you mean "it should take" or "realy takes" 16.9kW. It should take 0.9*17 = 15.3kW, and I cannot image it takes 16.9kWh.
In may case (the same setup as yours), to charge from 10% to 100% takes around 16kWh, so the efficiency is (shouldTake/realyTakes) 15.3/16=0.956

 

I read in a document posted that the onboard charger in the car is approximately 92 to 93% efficient. So it only loses about 7 to 8% to heat, the rest goes to the battery.
The most anyone has reported charging a 0 EV range, 2 battery bars, Clarity battery is 14.1 to 14.4 kWs. We don’t get to use the whole 17 kW because the BMS keeps a buffer at the top and bottom of the SOC in order to protect and prolong the life of the battery pack.
I think this is very difficult to track from the car side instead of from the wall side. Your best bet would be a “Killawatt” type device between the receptacle and the EVSE.

To track from the car side you might try using the % charge from the Honda Link app. Then approximate that 100% is 14.4 kWs and 0% is 0 kWs added. That’s my best guess until one is the EEs on the forum chimes in and then forget anything I said. Just remember that the Honda Link only updates by 15% increments so you have to let it go to completion or check it both before and after charging. So I guess if you start charging from say 25% and to to 100% then you added 75% of 14.4 kW or 10.8 kW. Then with a 93% efficiency or 7% loss, you really drew 11.55 kWs from the wall.
I hope I got that right. Everyone please check this.

 

I think that 100% SoC means that the battery is fully charged. If you your claim that it actually means "fully charged - buffer" to hide a "top buffer", than I would expect that at 0 miles EV it would show 0% SoC instead of 10% (two bars), to also hide a "bottom buffer".
I think that SoC displayed by Honda Link is the actual SoC, and that controller allows battery charge/discharge between 10% (or so) and 100%.

P.S. I am an EE, which doesn't mean that I am right (story of my life )

 

Why don't you buy a KillAWatt? You would remove the guesswork and have a way of resolving any disagreements about how many kwh you've used.

 

For $15 you can get a meter that will measure your usage so you don't have to think and will retain the info through power outages. I have been using mine since 2018. I pay my HOA quarterly and then reset it to 0.
https://www.amazon.com/gp/product/B00E945RMY/ref=ppx_yo_dt_b_asin_title_o07_s00?ie=UTF8&psc=1

 

I’m with you on the lack of electrical theory. Not my area of expertise either.

I think there has to be a buffer at the top of the SOC for two reasons. Battery health and to account for gradual loss of capacity. Charging Li-ion batteries to full capacity and discharging to full depletion are rather harsh on the chemistry.
Also, I’ve seen several posts that agree that the 2 bars in the bottom at 0 EV range is somewhat misleading or confusing compared to how it’s indicated at the top end. So in my scenario, I think you could use the Honda Link 0 to 100% as useful or allowable SOC that goes from 0% (2bars, 0 kW usable in EV only) to 100% (all bars, as fully charged as allowed). And even if it did charge with no buffer on the top end, the spread between 2 bars and all bars would still be ~14.4 since no one has managed to add more kW when charging from 0 EV and 2 bars.
OK, I quit, my head is hurting from trying to figure this out. It might have been easier for me just to buy a Killawatt for the OP!

 

Just last weekend, I charged from 10-99% SoC, and the Kill-a-Watt read 16.57kWh for that session. It definitely takes around 16.9kWh.
I rarely charge from 10% SoC, so the formula I use to predict energy needed for a full EV charge (keeping in mind that below 10% SoC is not usable for EV operation) is as follows:
Energy to charge = 90%*(Kill-a-Watt reading)/((SoC @ end of charge session)-(SoC @ beginning))

Hope that clears things up.

Also an EE here. Allow me to nerd out with you.

If you've ever messed with lithium batteries, you'll know that their behavior is nonlinear. They can be overcharged and over-discharged beyond safe capacity, and any off-the-shelf BMS will reserve buffers, some visible, some hidden. I can confirm for you that my 2018 Clarity reports 0.0 miles EV range, and will refuse to run off battery power once there are two bars on the dash, and the SoC readout is 10%, as reported through HondaLink and OBDII scanners. The space between 10% and 100% SoC is what I refer to as EV capacity.

Nominal battery capacity is larger than usable battery capacity, which is larger than permitted capacity, which is larger than EV capacity.
Based on pre-delivery inspections, the nominal capacity is around 55Ah@348V ≈ 19kWh. This would strongly suggest the 17 kWh advertised is whatever Honda's proprietary standards deem to be "usable."
The permitted capacity is the portion of the usable capacity which the BMS will allow the car to charge or drain before it forces the vehicle to shut off. This is represented by the % SoC, as read through HondaLink or OBDII.
The EV capacity is the portion of the permitted capacity which the computer will allow the motor to consume in EV mode before starting the genset to stabilize the charge. This range is 100-10% on the SoC readout. When most people refer to "usable" capacity, this is what they're talking about, because it's very difficult to discharge below 10% SoC.

We know the EV capacity is below 15 kWh, because users here report a charge from 10% to 100% SoC with a level 2 charger takes around 14-15 kWh from the wall. For me, it takes an average of 16.9 kWh. From this, I know that my wall-to-battery efficiency cannot be much higher than ≈85%.



Anyway, we're getting away from the subject matter. If you want to know how much power you're using, get a Kill-a-Watt. It's cheap and reliable. If you need a level 2 charger, get a JuiceBox or similar smart setup.

 

God bless you @Thevenin, that’s exactly the expert info we needed!!

 

I don't know specs of the battery in the Clarity, but one of the benefits of going with lithium batteries in renewable energy systems is their greater efficiency over other common types. Good quality lithium batteries are 98% efficient in conversion of electrical to chemical energy and back. Losses in the Clarity would include wiring, switching, connectors, etc., but it's hard to believe the engineering and materials involved would be more than a couple percent. I have and use a Killowatt device to measure usage in my solar car charging system, but I consider Hondalink to be such a coarse (really neither valid nor reliable in a statistical sense) estimate of SOC that I would not use the app to do any estimates of efficiency.

 

I believe ambient temperatures will be an important factor, as the internal resistance of the battery goes up quite a bit in cold conditions.

A 2013 Honda paper on the water cooled charging system of the Fit EV puts the on-board charger efficiency at 91.9% at 6.6 kW (slightly higher for 3 kW). Presumably the Clarity on-board charger is comparable.

 

Although it's several years old, I've seen this report used for efficiency data for L1 vs L2:

"On aggregate, Level 2 charging was confirmed to be 2.7% more efficient than Level 1 charging. A mean charge efficiency of 86.4% was observed in Level 2 charge events and Level 1 charge events were shown to be 83.7% efficient on average."

https://www.veic.org/documents/default-source/resources/reports/an-assessment-of-level-1-and-level-2-electric-vehicle-charging-efficiency.pdf

 

Here's a slightly more recent (2014) study in IEEE that looks at the same things:
https://ieeexplore.ieee.org/document/7046253

For the Chevy Volt and Nissan Leaf, it puts average efficiency of L2 @ 89.4% and L1 @ 83.8%. This disparity increases (87.2% vs. 74.2%) for charging sessions <4kWh. I can anecdotally confirm that some of my least efficient charging sessions are small top-offs, though nothing nearly that obvious.

 

Oops! That should be "I have been using my meter since 2014."