Prius FE Answers buried in the EPA Test Car Database

Thanks to Bob Wilson for directing me to http://www3.epa.gov/otaq/tcldata.htm

I started digging through the EPA Test Car Database and I think I have found some interesting data. First, I downloaded the test car data for each year 2001 through 2016. Then I stripped out all of the Prius Info and pasted it into a new spreadsheet. I also grabbed the Insight data to have a look at later. Since the data format changed between 2009 and 2010, I had to move the 2001 thru 2009 data into the appropriate 2010 thru 2016 columns.

I learned a lot. You can see the axle ratio and weight class changes over the generations. Gen 1 was tested in the 3125lb class. Gen 2 was in the 3250lb class. Gen 3 and Gen 4 is in the 3375lb class, except for the Gen 4 Eco, which drops down to the 3250lb class, same as Gen 2. I was also able to distinguish between the Base (Two, Three and Four) and Touring (Three and Four) Model tests. The key to that was the n/v ratio (engine speed versus vehicle speed at 50 mph). For 2016 models, the n/v ratio is 20.1 for the Touring models and 20.0 for the Base and Eco. For Gen 3, the n/v ratio is 26.9 for the Touring models and 26.8 for the Base models. The slightly higher number is due to the slightly smaller tire diameter of the 17” tires.

In addition to the normal FTP City and HWFET highway cycles, the 2016 data contain three additional tests. A cold weather version of the FTP City cycle, a SC03 City cycle with Air Conditioning use, and a US06 High Speed Highway cycle. The HWFET and the US06 have the same 48 MPH average, but the US06 reaches speeds of 80MPH were the HWFET only gets up to 60 MPH. All three City, FTP, FTP Cold and SC03 have 21 mph averages.

The raw FE data is of interest here. To make it more meaningful, I created a new column to normalize the data to the EPA estimates. I did that by applying a correction factor to the unadjusted FE of the base models that would make them equal to the EPA estimates. I used the same factor to normalize the Touring and Eco numbers to see how close they were to the actual EPA estimates. See the columns names RND_ADJ_FE, Normalized and EPA.

I have created two more columns to compare the results between models and generations. The 2016 Comp column compares all generations and models to the Base 2016 model. The 2001 Comp column compares all generations and models to the Base 2001 model.

Finally I have created the Load and Load % Columns. This is based on the target Dyno loading that was applied to each 2016 model during that particular test. The higher the load, the lower the MPG.

The Gen 3 models also had this data but it was not compatible with the 2016 loadings, so I left that data out so as not to confuse everyone.

Let’s start with the FTP City Cycle. You can see that the 2016 Base model was normalized to 100%, 54 MPG. Note that the ECO normalized to 58.2, slightly better than 58 EPA, while the Touring model normalized to 49.2,or 9% lower than the base model and 17% lower than the Eco model. It is clear that the Touring models are not included in the EPA 54/50/52 rating. Also note how when compared to the 2001 Base model, Gen 2 scores 16% better than Gen 1, Gen 3 scores 19% & 26% better than Gen 1 and Gen 4 scores 24%, 36% and 47% better than Gen 1.

I would also like to point out the Dyno loading. The Eco loading is 4% lower than the Base loading, 96% vs 100%, but the Eco has 8% higher FE. This leads me to believe that the weight class difference accounts for only half of the Eco/Base difference. What is the other half due to? The Touring loading is a whopping 22% higher than the base. That is the cost of “Bling” on the city cycle.



Next up is the HWFET Highway Cycle. The 2016 Base model was normalized to 100%, 50 MPG. The ECO normalized to 54.9, much better than the 53 EPA, while the Touring model normalized to 46.3, or 7% lower than the base model and 17% lower than the Eco model. Again, it is clear that the Touring models are not included in the EPA 54/50/52 rating. When compared to the 2001 Base model, Gen 2 scores 12% better than Gen 1, Gen 3 scores 15% & 21% better than Gen 1 and Gen 4 scores 14%, 23% and 35% better than Gen 1.

Eco loading is now 98% of base. Touring loading is 113% of base. The “Bling” penalty is much less on the Highway.



Next up is the US06 High Speed Highway Cycle. The 2016 Base model was normalized to 100%, 50 MPG. The ECO normalized to 53.1, slightly better than the 53 EPA, while the Touring model normalized to 46.8, or 6% lower than the base model and 10% lower than the Eco model. Again, it is clear that the Touring models are not included in the EPA 54/50/52 rating.

The loading for the US06 is that same as HWFET due to the same average speed. Eco loading is 98% of base. Touring loading is 113% of base.



Next is the Hot City Cycle, SC03 with Air Conditioning. The 2016 Base model was normalized to 100%, 54 MPG. The ECO normalized to 54.7, much worse than the 58 EPA, while the Touring model normalized to 47.6, or 12% lower than the base model and 13% lower than the Eco model. Ouch! That is a big hit for the Touring models. The “Bling” penalty is HUGE on the City Cycle with AC use.

The loading for the SC03 is that same as FTP due to the same average speed. Eco loading is 98% of base. Touring loading is 122% of base.



Finally the Cold City Cycle, Cold FTP. The 2016 Base model was normalized to 100%, 54 MPG. The ECO normalized to 55.7, much worse than the 58 EPA, while the Touring model normalized to 49.6, or 8% lower than the base model and 11% lower than the Eco model.

The loading for the Cold FTP is that same as FTP due to the same average speed. Eco loading is 96% of base. Touring loading is 122% of base.



Observations:

• If you opt for the touring models, you are going to see your FE 6% to 12% below the 54/50/52 Estimate for the base model, depending on driving conditions.
  
• If you opt for the Eco model, the 8% EPA advantage over the Base model holds up pretty well unless it is really cold, or really hot outside. In those cases, there is only a 3% and 1% advantage respectively.
• The mystery of the Eco model +8% over Base model may be only 50% explained by the lower weight class. What explains the other 50% is still unknown.

 

I can't answer that question without knowing more about how the weight class plays into the dyno loading, It may have no effect at all. It is really just my opinion at this point.

I understand that the coast down test is the basis for the dyno loading but it is not clear how the weight class effects the results. It seems like Toyota went through a lot of effort to get the weight class down one level. There must have been a reason.

You are right, I can't draw conclusions unless I have more facts. I have edited the text of my original post from conclusions, to observations. That way, others who are more informed can be the ones drawing conclusions.

 

I can't tell you what is correct, but I will tell you if I agree with your observations:

1. Agree
2. Disagree. After having discussed this very issue with another who has more knowledge about it that I do, I believe that the EPA is only interested in the high volume models and looks at the Touring models as outliers. As mentioned in the second paragraph of my original post, included in my definition of Base models are the two, three and four. This would account for the vast majority of Prius sales. Since the Eco is in a different weight class, separate testing was required which resulted in the 58/53/56 rating.
3. Disagree, see #2.
4. Agree, I don't know for sure about anything, just speculation on my part.
5. Agree that MPG for the touring is overstated significantly.
6. Agree
7. Agree, but Toyota is just taking advantage of an EPA rule.

Yes, I found the rule that lets all auto manufacturers do this. See section 7 paragraph 4 on page 4.

http://www3.epa.gov/fueleconomy/documents/420f14015.pdf

" Label regulations allow vehicles with the same engine, transmission and weight class to use the same fuel economy label value data, since, historically, such vehicle families achieve nearly identical fuel economy performance. Ford based the model year 2013 Ford C-Max label on testing of the related Ford Fusion hybrid, which has the same engine, transmission and test weight. For the vast majority of vehicles this approach would have yielded an appropriate label value for the car, but these new vehicles are more sensitive to small design differences than conventional vehicles because highly efficient vehicles use so little fuel."

Just now, after reading that quote, I am wondering if that was the reason the Toyota worked hard to get the Eco into a lower weight class, because they would have been forced to use the 54/50/52 label if it were in the same class. The rule that gives a higher than deserved rating to the Touring models would have given a lower than deserved rating to the Eco. Truly a double edged rule.

Disclaimer: Again, this is speculation on my part. I cannot know for sure what is correct.

It is nice to have another data set in which the same observations can be made.

It looks like the difference is reversed in the EU NEDC. I would have expected -6% on the Extra-Urban and -12% on the Urban cycle. Is that what you meant?

 

I wish that level of detail was in the data. I have not been able to find it if it is.

 

The 2016 Camry Hybrid LE has a weight equivalent of 3750. The Camry Hybrid XLE version has a weight equivalent of 3875. Weight classes seem to go in 125 lb increments.

Actually, there is a column labeled Test originator. Of the 42 tests listed, only 6 were initiated by the EPA. All were the base models (no Eco or Touring), two tests each on the FTP, HWFET and US06 test cycles. Not the SC03 or Cold FTP cycles. So my guess is that the hot and cold tests do not get factored into the EPA estimates. The remaining 36 tests were initiated by the manufacturer. Those included the Base, Touring and Eco models.

That would be my thought as well.

 

I really haven't studied the packages but If I were looking for ATP, I would get the three + ATP for 29k. Another thing to consider is the replacement cost of the tires. While I haven't prices 17's lately, I would bet they are at least 50% more costly than the 15's.