Predicting EV performance

[bwilson4web]

Hi,

Comparing EPA metrics for electric vehicles (EV) can be confusing. The EPA can publish 5-test metrics, 2-test metrics with an 'engineering factor', and/or manufacturer sandbagged value. For example, Tesla reports the 5-test metrics, the same as ICE and hybrid cars use, and is often criticized for 'too rosy' range. Fortunately, the EPA mandated, roll-down coefficients provide a single approach for head-to-head comparisons.

The roll-down coefficients measure the drag force as a function of speed:


The vehicle must supply an equal amount of thrust, the power, to maintain a constant speed:


These roll-down coefficients are for a "Standard Day". One could take an EV to a high altitude, hot summer day and make a benchmark with 'too good to be true' claims. For example, there was a Honda Insight team that did laps in the summer for bragging rights to an abnormally high range. More practical, cold weather can really hit performance such as my benchmark with a Gen 1, 2003 Prius, in freezing weather whose performance loss matched the density of the cold air. There are 'density altitude' techniques (and web pages) to predict these effects:


So here is an example of a spreadsheet comparing four EVs:


This approach unifies EV performance. Combined with the usable or available battery kWh, one can accurately predict range and electric operating costs. Yet is also works with ICE and hybrids.

The operating line and thermodynamic efficiency predicts the fuel consumption at any given speed. For example, the Toyota hybrid transmissions are very good about keeping the engine at peak thermodynamic range and banking or drawing excess power from the traction battery.

So I tried to attach a ZIP file with both this OpenOffice presentation and spreadsheet but the file size limit blocked uploading. Use this post for the presentation pages.

Bob Wilson
22-tstcar-2022-01-18_elect.xls

[aerohead]

At www.fueleconomy.gov, they give a breakdown of a BEV's energy bookkeeping:
* 10% charging loss
* 2% accessory loss
* 15% electric drive losses
* zero-to-2% auxiliary electrical losses
* 0% idle losses
* 6% Regenerative braking energy returned
* power to wheels = 77% - 79%
* overall 23% losses
------------------------------------------------------------------------------------
1) plus 47% aerodynamic loss
2) 23% rolling resistance loss
3) 7% braking loss
= 100%
--------------------------------------------------------------------------------------
If one has a strong bladder and minimizes rest stops, the significance of braking losses is diminished, allowing more available power for aero and rolling resistance.
Aerodynamic drag dominates the mathematics.
BEVs typically already have low R-R tires.
Aerodynamic drag reduction would be the obvious choice for range extending ( see the 2022 Mercedes-Benz EQXX concept ).

[bwilson4web]

Are these the URLs you meant to post:

• https://www.fueleconomy.gov/feg/atv-ev.shtml
• https://www.fueleconomy.gov/feg/evtech.shtml

I appreciate the references but I'm more intested in the gross vehicle performance. Based on the Munro reviews, there are significant differences in these first serious EVs. For example, some of them are significantly heavier than the more efficient EVs. Some use cheaper induction, more expensive switch reluctance, and other permanent magnet motors.

Think of my analysis as starting from the basic, high-level vehicle metrics and drilling down to more accurate details. Sad to say, the detailed specifics of each EV model are treated as proprietary. A generic description of parts is a little too imprecise for me.

I'm more interested in whether a vehicle is a 'Tesla killer' or just another also ran. The metrics you've described don't help in my quest.

Bob Wilson

[freebeard]

I'm more interested in whether a vehicle is a 'Tesla killer' or just another also ran. The metrics you've described don't help in my quest.

For me the number of wheels is top of mind.

[aerohead]

Originally Posted by bwilson4web

Are these the URLs you meant to post:

• https://www.fueleconomy.gov/feg/atv-ev.shtml
• https://www.fueleconomy.gov/feg/evtech.shtml

I appreciate the references but I'm more intested in the gross vehicle performance. Based on the Munro reviews, there are significant differences in these first serious EVs. For example, some of them are significantly heavier than the more efficient EVs. Some use cheaper induction, more expensive switch reluctance, and other permanent magnet motors.

Think of my analysis as starting from the basic, high-level vehicle metrics and drilling down to more accurate details. Sad to say, the detailed specifics of each EV model are treated as proprietary. A generic description of parts is a little too imprecise for me.

I'm more interested in whether a vehicle is a 'Tesla killer' or just another also ran. The metrics you've described don't help in my quest.

Bob Wilson

Neither are what I recall. Perhaps they revisit and 'freshen' data. Don't know.
I thought that perhaps you were interested in predicting what a specific vehicle might do if modified.
Also-rans can be modified to return Tesla-Killer-esque performance.

[redpoint5]

Seems like STP (standard temperature and pressure) applies here when making comparisons. Though I have never reviewed EPA testing criteria, I'm sure it would include this, with correction factors for testing outside of STP.

I wonder how long until the EPA busts the first overinflated EV efficiency claim?

[bwilson4web]

Originally Posted by redpoint5

Seems like STP (standard temperature and pressure) applies here when making comparisons. Though I have never reviewed EPA testing criteria, I'm sure it would include this, with correction factors for testing outside of STP.

The SAE sells the roll down specifications and procedure for $85. Over time, I've caught snippets and yes, it has fairly extensive Standard Day processes.

About five years ago, I tried to perform my own roll down tests using a recording GPS application and realized it would take more time and effort than I wanted to spend. The GPS altitude metric and delays of low cost GPS were the biggest problem. Combined with high accuracy accelerometer data, it might work.

Originally Posted by redpoint5

I wonder how long until the EPA busts the first overinflated EV efficiency claim?

The folks at "InSideEV.com" consistently claim the Tesla EPA metrics are too optimistic while other EVs vary. But their procedures leave a lot to be desired such as benchmarks at 70 mph with no density altitude corrections.

InSideEV reports the biggest difference is Tesla uses the 5-step procedure of the ICE world and other EV makers use a cheaper 2-step with an EPA mandated 'engineering factor.'

With over a decade of Prius experience, I'm a little persnickety about mileage and range metrics. Consumer Reports still pisses me off. Personally, I prefer the old Edmunds "Smackdown" series.

Bob Wilson