[2023 Prime] Watts, Watt-Meter, and Not What

I will be the first to admit I can get obsessive when trying to understand things. I want to measure everything, record everything, analyze and then analyze my measurements and analysis for errors and tolerance.

As many have suggested, the bottom line is cost per mile. When my new 2023 Prius Prime reports 33.9 miles driven at 3.8 miles per kWh, I start the analyzin'.

33.9 miles at 3.8 miles per kWh (with the AC on 75F with outside temp at 93F), took the SOC (State Of Charge) display from 100% down to 15%.

My first analysis was to see if these numbers sound reasonable. I have seen folks using 11 kWh as the available power for EV mode going from 100% to ----% (zero). So 85% of 11 kWh would be on the order of 9 kWh.

33.9 miles / 3.8 miles per kWh = 8.9 kWh - looks reasonable.

Next is to predict what using every available electron of battery power would allow under similar conditions:

33.9 miles / 0.85 = 39.9, so roughly 40 miles (and that is with the AC going full bore). Looks promising if it pans out.

Next to find out how much those 33.9 miles cost me. I hooked up the charger to my US$12 Chinese meter and started my stopwatch.

I also managed to get Dr Prius to successfully connect and found it reported the hybrid battery SOC at 32%. That also sounds reasonable, as even when the SOC showed ----% the color battery arc shows there is juice in the battery. This would suggest that reserve for HV mode to be around 17%.

Out of curiosity, checking the Toyota app showed battery at 47%. This does not sound reasonable and we'll try to understand more about this later.

8 hours and 15 minutes later, the Toyota app popped a notification on my phone that the charging was complete. I raced to the garage, with my keys, to disconnect the charger from the wall. ( I had determined leaving the charging cable / charger connected to the car and wall after charge completion will continue to draw power, and the watt-meter timer will continue to run as will the kWh accumulation advance slightly.

The meter showed 8 hours and 25 minutes but only 8 hours and 15 minutes had actually elapsed. (Confirmed by knowing the time I started charging, also confirmed by comparing the app "Charging Started" and "Charging Complete" notification times, and the stopwatch I started at the beginning.) Must not use the available 60 Hz events and is running fast. How fast? 2% and that just happens to correspond to one of the wattmeter's stated accuracies - 1% for voltage, 2% for current. They didn't spec the timer but now we know, 2% high. It might be possible the voltage is 1% low or high, the current 2% low or high, but the time is clearly 2% high. (I did a check with a 60W incandescent bulb - the meter showed 59W but I cannot know if the 59W is the real draw and the bulb is within some tolerance, or the meter power measurement is 1.7% low. If the meter power is 2% low and the meter timer is 2% high, it is possible the meter's kWh measurement is dead on. Considering the average daily charge is about $1 the 2% is probably 2 cents per day or 1/10th of a cent per mile. I should just ignore it but with time being 2% high, I'm going with the kWh 2% high.)

The meter showed 11.78 kWh drawn from the wall. Since I know the meter has used a 2% high time value in computing the 11.78 kWh, I choose to derate the wall power by 2% to 11.54 kWh.

The 120v AC had to be converted to higher voltage DC, and the cable resistance turned some of the current into heat, and the car was actively cooling the batteries in my 90degF garage, so it is reasonable to observe more power drawn from the wall to replace the 9 kWh used by driving. 11.54 wall / 8.9 battery = 1.3x wall to battery factor.

I have never seen the "cost" of this loss discussed. Just like using the MPG the car displays versus using the miles driven divided by the actual gallons, the miles per kWh can be the car value of miles per battery kWh or miles per wall kWh. In my case, miles per wall kWh is going to be 1/1.3 less or 23% less.

Car: 33.9 miles at 3.8 mi/kWh
Per adjusted wall power: 33.9 miles at 2.9 mi/kWh

Additionally, the cost per mile which is based on the cost of the wall power is increased:

11.54 kWh * $0.14294 = $1.65 and $0.049 per mile (real 2.9mi/kWh)

Comparing that to my three previous HV only mode fillups at $0.067 per mile (real 46 MPG)

Since the app would not have access to my power conversion factor, it could only report battery power of the charge. The app is totally crazy saying only 5.459 kWh was put into the battery to take it from the crazy 47% to 100%.

Also interesting is that Dr Prius says the fully charged SOC is 116.5%.

 

Or thought of as major failure to coordinate the software maintenance team with the model development team at any time during the final year the model is being readied for sale!

The charge kWh value is totally bogus, not in any way “convertible to make sense. I’ve tried linear regression, polynomial fit, cluster fit, and a even a back-prop neural net model and the best I can tell, it is using a wrong but more precise than displayed starting charge percent to compute a wrong delta charge percent times a wrong available battery capacity around 10 kWh. Admittedly I may be the only customer interested in the value, as even an accurate value is primarily useful as input to personalized equations.

The only correct app history information about a charge is the charge time (delayed availability after charge complete).

The car and app do a terrible estimate of “time to charge”. A better estimate of charging time with the provided Lvl 1 charger from a 120v 15A circuit, and the car changed to max charge (from default 8A):

miles driven / displayed average miles per kWh * 0.91 = hours

(Seems to charge battery at 1.1 kW until final 30 minutes at 0.6 kW)

and the estimated wall power will be:

miles driven / displayed average miles per kWh * 1.35 = kWh

(Wall rate is 120v * 12A = 1440 kW until final 30 min)

I’ve only done one short level 2 (free power!) charge so not enough data for modeling yet.

 

I drive with the current and average miles/kWh "Eco" window open next to the speedometer. Expose the multi-information display

Use the up/down, or left then up/down controls to select the "(leaf) Driving Info Display"

Then us the right/left control to display the Power Consumption Display:


In my 2009 Prius the indicated MPG (48.2) was 1.8 MPG high (46.4 actual). The energy in a battery, energy consumption, and recharge energy is much less knowable than physical gallons into a physical tank. Therefore I do not expect the indicated average power consumption (mi/kWh) to be as accurate as the indicated MPG, but it is a useful for estimating and modeling.