EPA Test Car Database with MPG vs mph charts

Introduction

In 2015, the EPA and Toyota conducted standard tests pre-production Gen-4 Prius and those results are found:
Test Car List Data Files | Cars and Light Trucks | US EPA

Thanks to a helpful e-mail and a few guesses, we've mapped the sales models to the EPA test data into a smaller table.

Test Car Database

This table lists the sales and tested Prius configurations versus the five EPA tests. The MPG numbers, "RND_ADJ_FE", are the lab tested, fuel efficiency, which factors out vehicle display errors. However, these numbers are further reduced to reflect owner experience and generate the Monroney sticker seen on new cars. So here is the data:

	sales\test	Cold CO	Federal fuel	HWFE	SC03	US06
1	Prius 2 ECO (Eco)	58.0(gas)	82.5(gas)	78.2(gas)	59.6(gas) 58.5(E10)	51.6(gas)
2	Prius 2 ECO (Nrm)	56.3(gas)	84.1(gas)	78.2(gas)	59.0(gas) 57.4(E10)	51.1(gas) 50.7(E10)
3	Prius 2 ECO (Pwr)	56.6(gas)	84.1(gas)	76.9(gas)	58.0(gas)	51.0(gas) 50.1(E10)
4	Prius 2 (NiMH 15")	56.3(gas)			58.8(gas) 58.1(E10)
5	Prius 2 (NiMH 17")	51.7(gas)			51.9(gas) 51.3(E10)
6	Prius 2 (NiMH all season tires)	55.4(gas)	77.3(gas)	76.1(gas)	52.3(gas)	48.6(gas)
7	Prius 2 (NiMH summer tires)	50.8(gas)	64.9(gas)	63.6(gas)	46.9(gas) 45.4(E10)	42.9(gas)
8	Prius 3_4 (15")		78.0(gas)	71.2(gas)		48.6(gas)
9	Prius 3_4 (17")		71.1(gas)	66.0(gas)		45.5(gas)

Each of three, Gen-4 Prius goes through up two five different tests: Cold CO, Federal fuel, HWFE, SC03, and US06. There are three Prius bodies tested: Level 2 ECO, Level 2 (NiMH), and combined Level 3 and 4 (hopefully a 4.) Then options are tested: different wheel sizes, standard and E10 gas, and two tire types. To get the best return on test time and labor, optional configurations are mixed to map out the boundary. This means some redundant tests, say "summer" versus "all season" tire configurations are skipped on some bodies. Still, these data raise eyebrows:

1. Prius 2 ECO (Eco), Federal fuel, 82.5(gas) - this data point is lower than the modes (Nrm) and (Pwr). It seems the test driver quality metrics, "DT-Inertial", "DT-Absolute", and "DT-Energy", are significantly different than the other two with both sign and magnitude differences. The (Eco) mode remaps the car response to the accelerator and this may have surprised the driver who has to follow a specific speed protocol.
2. "Prius 2 (NiMH 15")" and "Prius 2 (NiMH 17")" - were e-mail tagged as "Prius ECO" LiON models but their weight matches the "Prius 2". Also, "Prius 2" has the "summer tires" and "all season tires" tests which makes more sense for a compare and contrast than testing these tires on the "Prius 2 ECO." I could be wrong but IMHO this makes sense.

	sales\Federal fuel	RND_ADJ_FE		AVG BAGS	FE Bag 1	FE Bag 2	FE Bag 3	FE Bag 4		DT-Inertia Work Ratio Rating	DT-Absolute Speed Change Ratg	DT-Energy Economy Rating
1	Prius 2 ECO (Eco)	82.5		87.7	60.78	110.49	68.97	110.49		-0.47	-0.22	-0.35
2	Prius 2 ECO (Nrm)	84.1		90.7	61.24	110.52	66.44	124.53		0.08	0.15	-0.16
3	Prius 2 ECO (Pwr)	84.1		90.8	58.80	119.51	68.51	116.34		0.44	0.3	-0.18

1. Bag 1 and Bag 3 - start of each run, the cold MPG
2. Bag 2 and Bag 4 - end of teach run, the warmed MPG

This is the driving profile:

This test takes two days starting with a cold-start and capturing the exhaust in two large sample bags. Chemical analysis identifies the emissions but the bags also identify cold-start MPG and warm-start MPG. So we can map these results into the different Gen-4 sales models:

• Bottom markers - "morning" phases of a two day test. The car cools as if parked to replicate the first drive of the day. All of the exhaust gas is captured in bags, 1 and 3 on each day of the test and analyzed to determine emissions and fuel consumption.
  
• Top markers - "arrive at work" phase of each, two day test. A second bag, 2 and 4 captures the exhaust for emissions analysis.
  
• Warm-up phase - replicates "commuting" the car on a cross-town, commuter route. No measurements are taken.
• X-axis - the calculated, cumulative MPG from this test.
• Y-axis - the four, individual bag MPGs from this two day, test.

​

I've included the Gen-3 Prius, Prius v, Prius c, and PiP. The Gen-4 cold-start performance is much better, especially the Level 2 ECO. A lighter car, it is possible the traction battery is used more in this high-fuel burn region along with lower aerodynamic drag from the active vanes. However, warm-start performance is identical between the Level 3 and 4, 15" wheels, suggesting paying attention to warm-up driving profiles can bring the Level 3 and 4 close to Level 2 ECO performance.

The second take-away is tires and wheels have a significant effect. Just going from 15" to 17" wheels decreases mileage and here is the proof. They also tested a "summer tire" and it really hit mileage. It makes such a difference, one might consider writing the contract to replace the OEM tires with ones that also correct the in-car calibration error with lower rolling resistance.

EPA tests are defined so anyone can replicate and achieve similar results on a dynamometer. But dyno time is expensive and trying to follow the scripted driving protocol is not really practical for owners. However, we can do our own benchmarks mapping MPG as a function of speed, mph.

First we need a test route, 10-11 miles, without stops, without hills or undulations. On a straight, we do two runs in opposite directions and average the results to take out wind and altitude changes. A loop solves the two run problem because the start and stop is the same point. Loops are often found around shopping centers, industrial parks, and beltways around cities and towns.

The speed should be set and held by the cruise control. This takes out human errors and makes the data easily replicated. Since the minimum Prius cruise control speed is 23 mph and it can drop-out, start at 25 mph on the low side.

Conclusion

The Level 3 and 4, 15" wheel, Gen-4 has nearly identical performance to the Level 2 ECO. However, it may have safety options not available on the Level 2. Upgrade the tires and one should have Level 2 performance in a safer car.

Bob Wilson

ps. We have an answer via e-mail:

Hi Bob,

It looks like there is a bit of confusion about which vehicles are represented on the Test Car List. Hopefully this summary will clear up most of your questions.

Vehicle 16-ZV2H is the Prius 2 Eco, and it is tested in Power, Normal, and ECO modes (16-ZV2H/0, 16-ZV2H/1, and 16-ZV2H/2 respectively).

Vehicle 16-ZV1H is the Prius 3, 4 with the Li-Ion battery. 16-ZV1H/1 represents the models with 17 inch wheels, and 16-ZV1H/2 represents the models with 15 inch wheels.

Vehicle 16-ZV3H is the Prius 2 with the NiMH battery. 16-ZV3H/1 represents summer tires, and 16-ZV3H/2 represents all-season tires.

You can find descriptions of the sales model codes in the certification applications available on the Document Index System:
Basic Search | Document Index System | US EPA

Here are direct links to the applications:
https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=36706&flag=1
https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=36707&flag=1

The information is in the section “Description of Vehicles Covered by Certificate”

Regards,
<name withheld upon request>​

pps. I am working on a math model using the roll-down coefficients to map MPG vs mph.

 

I added more descriptions under the table and graph that hopefully explain what is going on. Sometimes I write in what reads like a foreign language, techno-babble. Equally obscure, mentally I often 'see' relationships between numbers which might not reach the text I post. So I am OK with suggestions to edit and improve.

Thanks to a helpful e-mail, most of the test vehicle models and configurations are mapped to the database. But the "summer" versus "all season" tires is unknown to me. Clever use of Google and the EPA search engine might find something describing what is going on. I could use some help in this area.

Bob Wilson

 

My eyes see the NiMH performing like the Level 3, 4.

This is an area where we could use some Google or search engine help. Re-read the release notes and perhaps someone can find something I've missed. Come on folks, team effort.

Bob Wilson

 

Thank you! This is exactly the type of dialog I am looking for:

This is exactly correct. The roll-down characteristics are significantly different between the Toyota and EPA coefficients. This is what we are talking about:

	Target Coef A (lbf)	Target Coef B (lbf/mph)	Target Coef C (lbf/mph**2)
1	17.130	0.28731	0.013857
2	18.272	0.29545	0.013793
3	19.034	0.31923	0.015397
4	20.302	0.32827	0.015326
5	24.153	0.33502	0.014096
6	26.837	0.37224	0.015662
7	31.145	0.35285	0.013956
8	34.605	0.39206	0.015507

Now the drag hp is:

drag_hp = mph * (Coef A + (mph * Coef B) + (mph**2 * Coef C) ) / 375​

But there is a factor of 2 between the lowest and highest Coef A which corresponds to the static (i.e., tire and weight) factor. But it isn't much better with Coef C, the primarily aerodynamic coefficient. So here is a graphical representation of the problem:



So the only remaining task is to calculate the gallons burned per hour per HP. Divide MPH by gallons burned per hour per HP and we have a set of curves showing the MPG as a function of speed. Fun with math.




We need some 10 mile, constant speed, benchmarks at 25 mph and the 60-75 mph range. Best case, a circular route with the same start and stop or two, bidirectional runs over the same, flat route. Include the model and tires; temperature; any calibration of odometer/speed against a known standard, and; any other factors that might affect the results.

The 25 mph benchmarks allow us to tweak the electrical overhead, estimated at 600W based on earlier Gen-3 metrics. The high speed benchmarks allow tweaking the drive train efficiency.

Bob Wilson