Lab vs ad hoc mileage testing, European

While researching another vehicle, I came across some credible, well written articles describing the difference between lab and ad hoc mileage tests:

http://www.carkeys.co.uk/features/technical/fuel_economy_testing_part_one.aspx

Fuel Economy Testing Part Two | Car Keys

Fuel Economy And Bad Science | Car Keys

I remain a consistent critic of Consumer Reports 'mileage' reports because we don't know enough about how they conduct their urban tests. They are consistently low suggesting the route and driving conditions are suspect. In contrast, the EPA and European tests are if nothing else, fairly reliable. To the extent that the new EPA protocols match real-world, owner reports, they have merit.

Bob Wilson

 

I appreciate the references but the CR protocol is explained as:

They did not describe the driving profile:

• acceleration profiles
• braking profiles

What I'm looking for could easily be provided with a time vs velocity chart similar to what we get from the Euro and EPA test series:

This would provide the details to understanding CR's concept of "City" driving.

Bob Wilson

 

I agree:

The USA tests were updated in 2008 to incorporate more highway and cold weather operation. You might find our EPA web site useful because the individual vehicle data also includes user mileage results.

I believe the EU test begins with a cold, urban cycle and then proceeds to the extra-urban cycle. Not to worry, it took me a while to understand the protocol and generate these charts:

Total EU testing profile with the 46 mph, 1.8L Prius threshold speed.

This is just the extra urban cycle that happens at the end:

The reason the extra urban cycle is important has to do with the threshold speed between 'hybrid mode' where the engine and traction battery cycle and 'engine on' where the engine runs all the time regardless of speed:

• 42 mph - NHW11 and NHW20, the 1.5L Prius
• 46 mph - ZVW30, the 1.8L Prius

Notice how the lower threshold speed of the 1.5L meant they were often just over the minimum, engine on, always time. But the 1.8L Prius is almost always running in hybrid mode where the ICE can shutdown. This has a practical effect on my 2003 Prius, NHW11 model.

I have replaced the front wheel tires:

• 175/65R14 - OEM tires, slightly undersized, they give higher indicated speed and overstate the mileage.
• 175/70R14 - rear tires, nearly right sized, they give slightly lower indicated speed and closer to true mileage.
• 195/70R14 - front tires, slightly oversized, they give ~7% lower indicated speed and worse indicated mileage. But I can stay in 'hybrid mode' up to 44 mph true speed which allows me to keep up with traffic on normal USA streets posted at 40 mph.

One of the articles referenced another that reports small engines with turbochargers reflects this design approach . . . to the test.

We must introduce you to:

CleanMPG Forums
<grins>

Understand I call myself an efficient driver, not a hypermiler, and I have very little interest in that 'art.' As an efficient driver, anything over EPA is not just 'good enough' but perfect (see signature.)

Bob Wilson

 

You raise a good question but I really haven't tried to understand how hybrids are 'normalized'. What is especially curious is how the Volt and other 'Plug-in' vehicles will be tested. But I wanted to share some aspects of the Prius battery management.

The Prius battery, 28 modules since 2004, are normally kept within an absolute range:

• 80% - maximum state of charge, any higher and it risks driving some modules into 'overcharge' and generating gas from electrolysis and excess heat.
• 60% - target state of charge, the hybrid computers will adjust charge and discharge to bring the traction battery to a long term, average of 60%. This provides 'head room' for braking and descending hills of ~160 m (525 ft.) at high speeds.
• 40% - lower limit to prevent weak modules from being too discharged.

We don't see 100% charge on the Prius traction batteries (nor the Ford hybrids) due to the risk to the traction battery life. This is just something many of us who have been in the 'hybrid' community often learn early. <wink>

Those are excellent article and something I follow closely. It is sad that too many of my fellow countrymen don't realize the limitations of tests and how they can distort any process being measured. But I was lucky in my Prius history.

The first week of October 2005, I took possession of our used, 2003 Prius. So even before I left Huntsville, I was recording, as is my practice, mileage:

Notice the "air miles" mileage in the upper right hand corner. I actually asked the flight crew about the amount of fuel burned on the two segments and calculated the end-to-end, air mileage. But I've done this since 1971 when I bought a 1966, VW MicroBus.

My 1500 cc, VW MicroBus had an air cooled engine with bolted on, cylinders and cylinder heads. It was common to find in service that the engine would 'stretch' and the valves would not close fully. This would lead to two effects: poor mileage, and; a 'pling' sound as the valves and valve seat gaps tried to 'torch' away. But I would see the mileage effect before the audio and would find a curb to back over and using a piece of cardboard to lay on, I would pop off the valve covers and adjust the valve backlash. Then the mileage would improve and life was good.

Later, I confirmed that any engine-only vehicle's mileage takes a hit when it needs maintenance:

• spark plugs
• oil change
• other, non-specific maintenance

So my first or second recommendation to anyone is "find a web site to record and track your mileage." So fast forward to October 2005.

My first leg from Fort Worth Texas to Shreveport Louisiana. On my drive across about 1/3d of Texas, I was driving like a thief who'd made a great deal. I kept up with Texas traffic, perhaps a little quicker, to reach the Louisiana border before they realized they'd screwed up and asked the Texas Rangers to 'send me back.' My average speed was a little over the posted 75 mph but I was keeping up with traffic. But my first leg, 39 MPG (6.03 L/100 km,) I was appalled and briefly thought about turnout around and returning the car. But then I realized that if the EPA can report at that time 45 MPG highway and 51 MPG city, I'll be a poor engineer if some government bureaucrat can beat my mileage!

Knowing the exponential effects of high speeds on aerodynamic drag, I drove the next segment at 60 mph. On that second leg, I got 51.7 MPG and knew better was possible. The third leg was at 65 mph, the posted night speed limit, and 51.2 MPG. I subsequently got ~49 MPG on a leg at 70 mph.

By happy accident, three weeks later I decided to record my mileage at GreenHybrid .com. This was before the "great schism" between the hypermiling advocates and efficient drivers. Understand, the only difference is as an efficient driver, I'm after "pass-fail," anything over the lab tested composite. I have modest performance standards suggesting that 50%+1 is not just good enough, it is perfect. But this moderation comes from my time at Greenhyrid .com.

I looked at the distribution of Prius mileage and noticed there were some folks who reported exceptionally great mileage in the NHW11 and others who report relatively low mileage, 40 MPG or lower. It was a Gaussian curve. What I learned is it was 'accident of home' that more than anything else set the end user's mileage.

I contacted one user who was a medical intern, a type of apprentice physician. He lived in Boston, a hilly city in a cold climate, at the bottom of a hill. He had killer hours (they often work 60+ hours per week and are 'on call') and when he needed to go to hospital, there were patients who needed his skills. He was reporting 39 MPG and I knew the Prius was providing the best, lowest cost, transportation available.

Another user lived at the top of a hill in North Carolina, a warmer climate than Boston. He was getting ~60 MPG until he started letting his son drive the car . . . hummm. In a flash of inspiration, I realized that to a greater or lessor extent, the accident of where someone lives and their 'real life' dictates their Prius mileage.

Now many of us know that there are techniques that tend to improve Prius mileage. If we can adjust our driving to deal with these characteristics, we can improve our mileage. But this takes study and an understanding of what different external and user managed effects have on mileage. So my studies continue.

Understand that I share my engineering insights with anyone and some of my Prius friends have feet in both camps. That is OK because my interest had nothing to do with ego, who gets most MPG. It has to do with my curiosity about life and an interest in sharing what little I've learned.

So I'm happy to see you have a 2009 Prius, the 1.5L or 1.8L version? The reason I ask is there are significant difference between the two. For example, the 1.8L does not require an extended warm-up but the 1.5L Prius thrives on block-heaters and gentle warm-up.

Bob Wilson

 

I agree although the biggest error seems to be their 'track' based 'City' driving. I know how to abuse a Prius on the track, Top Gear style, so that is the one I am most interested in seeing the velocity over time profile. But their highway numbers from the local interstate or divided highway are not so bad. If they would just publish the velocity profile on the 'flat' track, I could readily understand what they are doing.

Bob Wilson