EPA Methodology, is it accurately reflecting the actual driving conditions and resultant EV efficiency and range?

[hydbob]

I'd argue the EPA tests are the only reliable methodology of testing. The variables are controlled because they run the test on a dyno instead of in the real world. While real world conditions affect the actual efficiency, the EPA testing is the benchmark for a reason, because it's controlled. The consumers need to educate or be educated on what EPA testing means, not that it is flawed but that you need to adjust those ratings for your personal driving situation.

 

[momo3605]

Not sure why my post was deleted, but the answer really is "no." EPA methodologies will never accurately reflect real world driving because:

1. Too many variables (elevation, etc)
2. Differing driving habits between individuals
3. The fact that the EPA cannot test at speeds such as 75 mph, as they likely legally need to test at the speed limit or below.

No matter how the methods are changed, the EPA will never be 100 percent accurate. What they can do is get closer to reality, and they also should put one unified testing standard in plase.

The answer is NO, but not because of the reasons you suggest.

The 5-cycle EPA test has an adjustment factor. Manufacturers are free to bullshit their way into higher numbers. Tesla, Rivian, and Lucid all use the 5-cycle test, but even if you put the same variables, same speed, same elevation, same temperature. Lucid always falls short, Rivian always beats. It's because of the adjustment factor.

There are 2 answers. EPA stops with the adjustment factor nonsense on the 5-cycle test, OR Lucid stops using an aggressive factor and uses something that aligns more with real-world testing
The Adjustment Factor Tesla Uses to Get Its Big EPA Range Numbers (caranddriver.com)

I'd argue the EPA tests are the only reliable methodology of testing. The variables are controlled because they run the test on a dyno instead of in the real world. While real world conditions affect the actual efficiency, the EPA testing is the benchmark for a reason, because it's controlled. The consumers need to educate or be educated on what EPA testing means, not that it is flawed but that you need to adjust those ratings for your personal driving situation.

Again, same answer to you...read above, the EPA test is not valid because it's not controlled. Out of Spec and Edmunds and C&D range tests are more controlled than the EPA test

 

[BS8899]

I'd argue the EPA tests are the only reliable methodology of testing. The variables are controlled because they run the test on a dyno instead of in the real world. While real world conditions affect the actual efficiency, the EPA testing is the benchmark for a reason, because it's controlled. The consumers need to educate or be educated on what EPA testing means, not that it is flawed but that you need to adjust those ratings for your personal driving situation.

I don't think one can replicate every conceivable modulating factor into one test.

That said, touting the EPA mixed city/hwy (55 mph?) number as the "range" of the vehicle is, at best, misleading. Yes, it might fit the "legal" definition. But, as many of us noted, the range number is most important on highway/intercity drives. As such, the current instantiation of what we call an EV's range is misleading.

At a minimum, if we have a highway efficiency and range number, that would make it more useful.

To @hydbob's point, if EPA testing is a car in a static environment on a dyno, is it even meaningful? When we talk about wind, drag factors, etc., when you drive on the highway, it is not exactly "laminar-flow" over the car. Hence, your drag factor is not 0.2!

having a set of lab-tested data (say EPA) is a benchmark. But touting or even implying it as range is ludicrous!

 

[hydbob]

I don't think one can replicate every conceivable modulating factor into one test.

That said, touting the EPA mixed city/hwy (55 mph?) number as the "range" of the vehicle is, at best, misleading. Yes, it might fit the "legal" definition. But, as many of us noted, the range number is most important on highway/intercity drives. As such, the current instantiation of what we call an EV's range is misleading.

At a minimum, if we have a highway efficiency and range number, that would make it more useful.

To @hydbob's point, if EPA testing is a car in a static environment on a dyno, is it even meaningful? When we talk about wind, drag factors, etc., when you drive on the highway, it is not exactly "laminar-flow" over the car. Hence, your drag factor is not 0.2!

having a set of lab-tested data (say EPA) is a benchmark. But touting or even implying it as range is ludicrous!

But that's the point of a benchmark. Everyone can refer back to it and then adjust for their own environment. I don't believe it is the role of the manufacturer, ICE or EV to account for everyone's personal use case, which is why you rely on the EPA benchmark as a starting point and adjust off that number.

 

[BS8899]

I don't think one can replicate every conceivable modulating factor into one test.

That said, touting the EPA mixed city/hwy (55 mph?) number as the "range" of the vehicle is, at best, misleading. Yes, it might fit the "legal" definition. But, as many of us noted, the range number is most important on highway/intercity drives. As such, the current instantiation of what we call an EV's range is misleading.

At a minimum, if we have a highway efficiency and range number, that would make it more useful.

To @hydbob's point, if EPA testing is a car in a static environment on a dyno, is it even meaningful? When we talk about wind, drag factors, etc., when you drive on the highway, it is not exactly "laminar-flow" over the car. Hence, your drag factor is not 0.2!

having a set of lab-tested data (say EPA) is a benchmark. But touting or even implying it as range is ludicrous!

Real-life range testers such as Kyle (OoS) might not be perfect. But their data is a quantum leap more applicable to real-life.

We can bicker as to Kyle's testing in CO, elevation, temperature/road conditions, etc. till the cows come home. I am sure Kyle is aware of these nuances. But his data is far closer to reality than what most manufacturers tout.

 

[Adnillien]

The answer is NO, but not because of the reasons you suggest.

The 5-cycle EPA test has an adjustment factor. Manufacturers are free to bullshit their way into higher numbers. Tesla, Rivian, and Lucid all use the 5-cycle test, but even if you put the same variables, same speed, same elevation, same temperature. Lucid always falls short, Rivian always beats. It's because of the adjustment factor.

There are 2 answers. EPA stops with the adjustment factor nonsense on the 5-cycle test, OR Lucid stops using an aggressive factor and uses something that aligns more with real-world testing
The Adjustment Factor Tesla Uses to Get Its Big EPA Range Numbers (caranddriver.com)



Again, same answer to you...read above, the EPA test is not valid because it's not controlled. Out of Spec and Edmunds and C&D range tests are more controlled than the EPA test

The adjustment factor used in the 5-cylce tests is determined by the three additional cycles. It is not a number made up by each automaker but an adjustment number determined by the EPA methodology. The EPA methodologies, there are two, are standard and repeatable but it would be better if there were only one methodology.

 

[xponents]

The adjustment factor used in the 5-cylce tests is determined by the three additional cycles. It is not a number made up by each automaker but an adjustment number determined by the EPA methodology. The EPA methodologies, there are two, are standard and repeatable but it would be better if there were only one methodology.

One question for you regarding this; I thought I read an article before stating that Tesla petitioned the EPA for a higher adjustment factor? Did Rivian purposefully adjust their factor lower, or did they just "get" the lower factor? And what does the "methodology" that determines the factor involve?

 

[hydbob]

The answer is NO, but not because of the reasons you suggest.

The 5-cycle EPA test has an adjustment factor. Manufacturers are free to bullshit their way into higher numbers. Tesla, Rivian, and Lucid all use the 5-cycle test, but even if you put the same variables, same speed, same elevation, same temperature. Lucid always falls short, Rivian always beats. It's because of the adjustment factor.

There are 2 answers. EPA stops with the adjustment factor nonsense on the 5-cycle test, OR Lucid stops using an aggressive factor and uses something that aligns more with real-world testing
The Adjustment Factor Tesla Uses to Get Its Big EPA Range Numbers (caranddriver.com)



Again, same answer to you...read above, the EPA test is not valid because it's not controlled. Out of Spec and Edmunds and C&D range tests are more controlled than the EPA test

How is testing in a lab and applying a set factor to the results LESS controlled than getting on a road and driving? The EPA tests might not result in real life metrics, but there no way that lab testing is LESS controlled than someone driving on a road.

 

[BS8899]

How is testing in a lab and applying a set factor to the results LESS controlled than getting on a road and driving? The EPA tests might not result in real life metrics, but there no way that lab testing is LESS controlled than someone driving on a road.

Let's ask the question differently:

> EPA is a lab test in a controlled environment.
> given exactly the SAME INPUTS (adjustment factors) and test condition, are the results reproducible?

I think the answer is probably YES.
This is what some manufacturers are latching onto.
But the pertinent question is:

 

[BS8899]

Let's ask the question differently:

> EPA is a lab test in a controlled environment.
> given exactly the SAME INPUTS (adjustment factors) and test condition, are the results reproducible?

I think the answer is probably YES.
This is what some manufacturers are latching onto.
But the pertinent question is:

But the pertinent question is: on highway driving (mph, roads, typical wind etc.) where range and efficacy matters, does the EPA number tell us what we need to know?

I think all the EV manufacturers know the real answer. These manufacturers have all the data we have and then much more. Many are just playing the game by answering the question "Yes, we meet the EPG metrics". It becomes a self-consistent non-answer.

Walter Scott's famous quote comes to mind:

"O, what a tangled web we weaved, when first we practice....."

 

[Buffalo Bob]

So, I thought it might be fun to list those factors that would potentially affect actual range vs. EPA range one way or another. Lots of stuff that I bet affects ICE cars in similar ways, but it doesn't hit people's radar screens when a gas station is around every corner.

I bet I'm missing some...

Speed at cruise during both city & highway driving.
City/Highway driving mix.
Rate of acceleration & deceleration.
Use/Level of regenerative braking.
Outside temperature.
Battery temperature.
Use of interior heating or cooling.
Use of lights.
Tire selection.
Tire inflation.
Use of aero covers.
Wind vs. dead calm. Headwind/Downwind/Crosswind.
Loading of the vehicle.
All windows closed.
Precipitation.
Test area altitude above sea level.
Elevation increases/decreases vs. flat.
Being chased by the police. OK. Not that one.

What else?

 

[Chizle]

So, I thought it might be fun to list those factors that would potentially affect actual range vs. EPA range one way or another. Lots of stuff that I bet affects ICE cars in similar ways, but it doesn't hit people's radar screens when a gas station is around every corner.

I bet I'm missing some...

Speed at cruise during both city & highway driving.
City/Highway driving mix.
Rate of acceleration & deceleration.
Use/Level of regenerative braking.
Outside temperature.
Battery temperature.
Use of interior heating or cooling.
Use of lights.
Tire selection.
Tire inflation.
Use of aero covers.
Wind vs. dead calm. Headwind/Downwind/Crosswind.
Loading of the vehicle.
All windows closed.
Precipitation.
Test area altitude above sea level.
Elevation increases/decreases vs. flat.
Being chased by the police. OK. Not that one.

What else?

Paint color. Red makes cars go faster.

 

[Buffalo Bob]

Paint color. Red makes cars go faster.

How did I miss that one? Oh, I know... It's because my car is 'Slow-Assed Silver'.

 

[xponents]

Paint color. Red makes cars go faster.

I mean, it also increases your chance of being chased by the police, so in the end, Boring Silver would be the fastest color!

 

[Adnillien]

So, I thought it might be fun to list those factors that would potentially affect actual range vs. EPA range one way or another. Lots of stuff that I bet affects ICE cars in similar ways, but it doesn't hit people's radar screens when a gas station is around every corner.

I bet I'm missing some...

Speed at cruise during both city & highway driving.
City/Highway driving mix.
Rate of acceleration & deceleration.
Use/Level of regenerative braking.
Outside temperature.
Battery temperature.
Use of interior heating or cooling.
Use of lights.
Tire selection.
Tire inflation.
Use of aero covers.
Wind vs. dead calm. Headwind/Downwind/Crosswind.
Loading of the vehicle.
All windows closed.
Precipitation.
Test area altitude above sea level.
Elevation increases/decreases vs. flat.
Being chased by the police. OK. Not that one.

What else?

I you really want to get into the weeds, barometric pressure can be added.

 

[HC_79]

Being chased by the police. OK. Not that one.

Clearly you're not from LA

 

[Unmutual]

cleanliness of vehicle. mood of spouse. type of music (thrash vs. classical)

 

[hydbob]

So, I thought it might be fun to list those factors that would potentially affect actual range vs. EPA range one way or another. Lots of stuff that I bet affects ICE cars in similar ways, but it doesn't hit people's radar screens when a gas station is around every corner.

I bet I'm missing some...

Speed at cruise during both city & highway driving.
City/Highway driving mix.
Rate of acceleration & deceleration.
Use/Level of regenerative braking.
Outside temperature.
Battery temperature.
Use of interior heating or cooling.
Use of lights.
Tire selection.
Tire inflation.
Use of aero covers.
Wind vs. dead calm. Headwind/Downwind/Crosswind.
Loading of the vehicle.
All windows closed.
Precipitation.
Test area altitude above sea level.
Elevation increases/decreases vs. flat.
Being chased by the police. OK. Not that one.

What else?

Tire tread life also affects range

 

[Buffalo Bob]

Gee, I thought I'd be missing some glaring ones, but so far, I'm not too far off the mark.

Maybe we can get Billy Joel to recreate an EV version of 'We Didn't Start The Fire' with all of these listed.

 

[borski]

The answer is NO, but not because of the reasons you suggest.

The 5-cycle EPA test has an adjustment factor. Manufacturers are free to bullshit their way into higher numbers. Tesla, Rivian, and Lucid all use the 5-cycle test, but even if you put the same variables, same speed, same elevation, same temperature. Lucid always falls short, Rivian always beats. It's because of the adjustment factor.

There are 2 answers. EPA stops with the adjustment factor nonsense on the 5-cycle test, OR Lucid stops using an aggressive factor and uses something that aligns more with real-world testing
The Adjustment Factor Tesla Uses to Get Its Big EPA Range Numbers (caranddriver.com)

That is literally not true. From the exact same article you linked: “The default adjustment factor reduces the window-sticker range by 30 percent. So a car that achieves 300 miles of range during the city-cycle dynamometer test ends up with a 210-mile city rating. However, the EPA allows automakers the option to run three additional drive cycles and use those results to earn a more favorable adjustment factor.”

Neither Lucid nor Tesla “comes up with” their adjustment factor. It is calculated based on the additional three cycles on top of the basic two. Anyone can choose to do those extra 3 cycles. There is no conspiracy here, and there is no lying. Nobody is “fudging numbers.” Quit trying to make fetch happen, lol.

Also from the same article:
“This dual-method testing dates back to the EPA's 2008 overhaul of fuel-economy labels. Initially, the agency wanted all vehicles to undergo five-cycle testing, but automakers claimed it was an overwhelming burden and the EPA relented. The agency then came up with a way to convert two-cycle results into five-cycle ones, and today, many automakers exploit this gamesmanship between the two methods for their gas-powered vehicles.

When it comes to EVs, though, the EPA came up with the default 30 percent adjustment factor back in 2011, when there were very few electrics on the market to analyze. The agency tells us it initially plucked the figure from the test results of the Toyota Prius, and after a subsequent analysis in 2015 showed that the Prius's factor was a solid guess, it chose not to change it.”

So it isn’t just EVs; gas cars use the same style of cycles, it’s just that nobody actually thinks about the range of a gas car because there are so many gas stations and they’re so quick.

As usual, the people you want to blame here are your federal government, and you should write to your government to change the EPA testing guidelines.

But it isn’t Lucid or Tesla or Rivian or BMW or Hyundai or anyone else lying. It *does* happen sometimes, because companies want to get the best numbers they can, but they get caught and smacked. Also from your linked article: “Hyundai got in trouble for this in 2012. The company used highly optimized tires and cherry-picked favorable results that turned out to be unachievable in the real world. It got hit with a $100 million fine from the EPA in 2014 and reduced fuel-economy ratings across much of the 2012 and 2013 Hyundai and Kia lineups.”

Now… can we please, please, please, please, please… put this to rest?