Power used for charging compared to power stored in the battery

[twiggy144]

I have a power meter (I had bought off ebay) to measure the power used by several domestic appliances. It measures the voltage, the current in amps, the instant power ( volts x amps) and accumulates the power consumed in KWH . I validated the precision of this power meter in various conditions, with various appliances, and I can vouch it is very accurate, with a +/- 5% reading error.

I plugged this power meter between my house wall outlet and my 2025 Outlander's 120 Volt charging cable, and I recorded the kwh for each recharge over a period of three weeks, or approximately 10 recharging cycles. Here are my findings so far.

The average ambiant temperature varied between 0C and -12C for this observation period.

The voltage at my power outlet varied between 115 V and 121 Volts.

Before and after each recharge, I recorded the battery state of charge ( SOC) from the dashboard in sixteenths, from zero to 16 sixteenths, corresponding to the number of bars displayed in the battery SOC meter.

When the battery SOC is between zero and 14/16 , the charging amperage is stable between 10.9 and 11.1 amps.

Once the battery SOC exceeds 14/16, the charging amperage reduces to 3.5 amps.

The accumulated power measured by the power meter for a full charge from zero SOC to 16/16 is 20 KWH, whis is exactly the capacity of the battery as claimed by Mitsubishi.

I observed the accumulated power measured by the power meter while the battery SOC was gradually climbing from 1/16 to 14/16, it matches exactly the power accumulated as displayed in the SOC, given each 1/16 is equivalent to 1.25 kwh ( 20 kwh divided by 16 = 1.25 kwh). This means the time lapse recharging curve is linear (directly proportionnal) between zero and 14/16 SOC.

Time for a full recharge varies between 16 and 17 hours, which matches the full recharging time claimed by Mitsubishi ( 16 hours).

Because my power meter measured 20 kwh used for a full recharge of the 20kwh battery, (with an error of +/- 5% reading on my power meter), there appears to be a negligeable loss of power in heat during the recharging process.

I observed the km range as displayed by the guess-o-meter during my experiments. The estimated range displayed is way too wild to estimate the energy stored in the battery. We all now know the guess-o-meter varies way too much and uses too many varying parameters such as HVAC operation, seat heaters, driving mode, etc. Fortunately, at least the battery SOC displayed by the number of bars between 1 and 16 reflects accurately the energy stored in the battery as per my findings.


Comments?

 

[twiggy144]

I have an adjustment to make. Today I charged with a SOC nearly zero, for 4 hours , on 114V with 10.9 amps. Total energy used was 4970 wh.
The battery SOC should have indicated between 4/16 and 5/16. It displayed 9/16, double the expected reading. Dont know why. I will keep obeerving and recording data.

 

[twiggy144]

I did a little research. Found this site;

https://evreporter.com/understanding-charge-discharge-curves-of-li-ion-cells/

The site looks serious with valid information. It has the following graph reproduced below.

From what I understand, when the total practical charging time is 2.0 hours for 98% charge, after 0.5 hours, the charge capacity had already reached 50%. So after 25% of the expected charging time, the battery would have been recharged to 50% ? Would this explain why in my above post I had observed the battery SOC was double of what I was expecting after 4 hours of recharging?






Again , the explainations from the site reads:




This means even if in the dashboard the battery SOC indicates close to nearing full charge, it does not mean that it is 100% charged.


Any comments?

 

[AndyInOz]

Other forum members have mentioned that they see this behaviour when monitoring charging.

 

[PHEV07]

Thanks for this great information.
I will try to compare this info with my 2024, once I get it back from the Dealer and the ICE Remote Starting has been installed

 

[mountkay]

As someone with a lot of experience with solar and lithium iron phosphate batteries, I can tell you that measuring those types of batteries SOC based purely off of voltage is very hard. If you look at the chart you posted, the difference in voltage between 30% SOC and 100% SOC is about 0.2v per cell.

I’m not sure what type of lithium-ion chemistry is used in the Mitsubishi batteries.

I’m also not sure how Mitsubishi measures the SoC of their battery packs (is it based off of voltage or do they use a shunt?)

People that want an accurate reading of their solar LFP use a Shunt that reads the actual amperage entering and leaving their battery packs because simply using voltage is very inaccurate.

This means even if in the dashboard the battery SOC indicates close to nearing full charge, it does not mean that it is 100% charged.

The most accurate way to tell if the at battery is full would be to see how much power the charger is pulling. Like the explanation you posted, once the cells reach full voltage, the amperage will start to drop, and the battery will be “trickle charged”.

You observed this when you noticed the chargers amperage drop to 3.5 amps. My charger will use its full power and then slowly drop its power when the bulk charging is done. It will steadily lower 900w 800w etc. 500w, 300w, 200w, 100w etc until it requires basically no power to keep the cells at their full voltage.

A regular lead acid battery follows the same charge curve, first you have full amperage (constant amperage) of the charger. Then when full voltage is reached, the amperage will slowly lower until it requires only 1a to keep the cells at their full voltage (~14.6v for lead acid).

Lead acid batteries are much easier to read their SOC off of voltage tho, unlike lithium iron phosphate.

 

[mountkay]

Here is the difference of SoC when read based off of voltage for a lithium ion battery vs lead acid. You can see a lead acid has a more direct relation between voltage and SOC. Whereas lithium ion has a much flatter curve


Here is a basic chart showing voltage to SOC of a 12v lithium ion battery. As you can see, the difference of 70% - 40% is only 0.1v.

Hopefully this all shows you how reading SOC off of lithium batteries is very hard if you are doing it strictly off of voltage.

Again, im not sure how Mitsubishi measured the batteries SOC, but it’s possible this could account for the variation in SOC that is being displayed to you

 

[twiggy144]

So, I gave up relying on the battery SOC ( numbers of bars from 0 to 16 displayed in the dash) to estimate the electric energy consumed. I decided on using the following technique instead.

I read the cumulative kwh from my power meter, which is plugged between my wall outlet, and the vehicule's charger. I compare this reading, with the kwh / 100 km reading displayed in the dash.
Here are my results so far:

km driven: 453km, kwh accumulated by the power meter: 97 kwh, for a rate of 97/453 = 21.4 kwh / 100 km.

The vehicule's computer displays 24.3 kwh / 100 km for the same distance of 453 km.
The difference between both readings is 8.8%.

Analysis:

First of all, I say I have complete trust in the accuracy of my power meter, for having it tested on numerous different electrical appliances in my home. So I take the reading of 97 kwh, hence the rate of 21.4 kwh / 100 km, for granted.

The vehicule's computer displays 24.3 kwh / 100 km, or 8.8 % more than actual . Why this gap? I can offer the following explanation. On one of my rides, after switching off the engine, I noticed the dash briefly displayed I had gained 4 km with regenerative braking. This was for a 50 km ride. The gain in % is 4km/50km = 8% I dont know if this figure is average regeneration, but it matches the 8.8% gap in my observations.

If this explanation makes sense, it would also mean the energy used displayed by the vehicule's computer does not take into account the energy generated by braking. It would measure the energy coming out of the battery, and would not deduct the energy regenerated.

Comments?

 

[nev..]

Analysis:

First of all, I say I have complete trust in the accuracy of my power meter, for having it tested on numerous different electrical appliances in my home. So I take the reading of 97 kwh, hence the rate of 21.4 kwh / 100 km, for granted.

The vehicule's computer displays 24.3 kwh / 100 km, or 8.8 % more than actual . Why this gap?

This seems counter intuitive. Because of losses involved in the charging process, such as heat losses in the EVSE, the charging cable and the traction battery, I would expect the power meter at the wall socket to measure a higher usage than the vehicles internal meter (because it's going to push more than 20kWh of power into the battery to get that battery to 20kWh capacity, and therefore the power meter would have a higher calculate kWh/100km rate than that displayed on the vehicle dash.

I have been keeping my own record of power consumption in a similar fashion since I bought my 2024 ZM Phev, and the display (for all 3 trip meters on the dashboard shows between 20.5 and 20.8 kWh /100km while my calculation from the power meter at my wall plug equates to about 24kWh/100km. This seems reasonable to me, although maybe a bit low on the dashboard trip computers. A few times on hot days I have used the climate timer to run the AC to cool the car before I leave work, so it seems reasonable that this battery usage while the traction drive systems are shut down would not be included in the dashboard calculations. I'd be interested to see other people's calculations.

 

[twiggy144]

I agree nev, counter intuitive. I would expect the power as measured by the power meter to be higher.
I will keep monitoring.

 

[littlescrote]

Can we please use power and energy correctly?

Power is instantaneous (kW)

Energy is cumulative over time (kWh).

Please don't mix the 2.