Prius overall MPG shows no improvement in the 3rd generation according to Consumer Reports tests

I'm not a statistician, but I think using the same pump increases your error, long term. When averaging over many tanks there are only two things that count.

First, is your very first and very last tank levels the same. The more tank fulls you use the less and less this matters. Perhaps you could argue that your first and last fillups should be at the same pump.

Second, what matters is how accurately the reported gallons are for each fillup. Using just one pump causes you to always have the error for that pump. Using many pumps gets you some pumps that have more or less error, again many fillups will average you out to whatever the common error level is, not the exact error of a single pump.

By law (probably the same in every county/state) the pumps are supposed to be accurate plus or minus 0.5% on a 5 gallon fillup (maybe 10 ?) I assume that there is a bias in that they are all more on the minus side than the plus side. But if this wasn't the case you'd sometimes get more than the indicated amount and sometimes less...thus over many fillups you'd get very close to your actual recorded amounts.

3PriusMike

 

I do not think this is completely correct and tends to propogate the (IMO) myth that the Prius gets higher mpg in the city because the engine shuts off at stops or because of regenerative braking. Both of these actions reduce the losses that would have occured otherwise. These losses do not even happen in highway driving (assuming no traffic, etc). They do not increase the mileage that you would get if you just drove at city speeds and never stopped. The fact is that you get better mpg at slower speeds. AND at slower speeds, under right conditions, the ICE can shut off, then run later at a more efficient RPM and charge the battery, go uphill, accelerate, etc.

3PriusMike

 

My gas receipts show three significant digits. Unless that last digit is noise, accuracy is quite a bit more than 1/200, and does not leave room for bias that would matter.

Addendum: Found this through Google

A gallon is 231 inches cubed, so the inaccuracy of a 10 gallon fill-up is about 1/400. There is a lot more inaccuracy in the distance meter.


​

 

Actually the dynomometer tests are more repeatable and they do take into account tire rolling resistance (the tires are on rollers that are "simulating" a fairly smooth road) and the weight and Cd are used to mathematically adjust the numbers measured to lower mpg values.

3PriusMike

 

I believe that CR does a fair job in rating the driving characteristics of each vehicle. I did see some values from an article somewhere previously that rated a lot of V6 sedans in the low teens for their city driving segment whereas the EPA values were in the low 20's.

I agree that as long as the protocols are the same in every test then all the results are 'valid'.

I had a long discussion on Edmunds with someone who I think was involved with CR and its testing. In the end he acknowledged that on their highway run the Prius G2 got 48 mpg. However on their city run it only got in the low 40's. Their city run consisted of a lot of stops with a lot of short trips over their own course.

Rightly or wrongly I think that CR goes out of its way to point out warts in order to protect the buying public. Many won't agree here but this is their mission and I think that they accomplish their goals. Screw everybody else.

That being said, I never read CR for any reason.

 

Somebody is buying somebody off... thats silly... I'm getting an easy 5mpg better.. sometimes ranging to 10mpg better!

And I'm comparing a brand new car thats not even broke in yet and hasn't even had one tank of gas ran all the way through it with my other prius's that have 91K on them each.

 

The number of significant digits represents precision, not accuracy. In engineering, it is very important to not confuse these concepts.

 

I'm saying they need a longer test to incorporate warm-up of the transmission and tires plus some way to always start the car at the same starting temperature. If they simply garage the car at say 60-70F, I would be happy that the 'cold soak' problem was resolved. Their report says they 'adjust for temperature' and that is typically done to handle air density. The vehicle 'cold soak' is what remains unknown. For example, this is the warm-up profile for my NHW11 transaxle:

You'll notice it takes at least 20 minutes on a warm day to reach a stable transaxle temperature. Yes, transaxle temperature is important because it impacts energy losses from engine-to-wheel:

One might as well ride the brakes after 30 minutes to replicate the highway, must less their City performance.

I saw nothing to worry about in their handling wind effects and traffic should not be anywhere near the vehicles during the urban testing. I'm happy with that part of their protocol and actually like it.

It is easy to adjust for air density but things like engine and transaxle heat losses are more difficult, especially when temperatures are under 50F and it gets wicked at under 40F. For example, looking at mileage vs temperature from the old GreenHybrid database:

Certainly the dyno makes sense in getting reproducible results and I wish I had one. I've found a local speed shop where I've thought about doing some tests. However, when I figured out how to use MG1 torque to measure ICE torque, my interest faded. Still, I keep thinking about making a portable, dyno to take to hybridfest. Then I think about the liability and look at other projects.

Bob Wilson

 

We call that "sand bagging" and if we realize that is what they are doing, no problem. Off hand, it looks like they are striving for about one standard deviation down in typical vehicle performance. However, if one is too aggressive in "low balling" the results, it tends to cluster everyone 'in the dirt.' One thing I don't want to do is get into a detailed analysis of the different testing protocols.

In a perfect world, every car would have a data recorder and wireless communications link to every fuel pump. When we filled up, the pump would also use the vehicle data and then add to the receipt suggestions like:

• mileage is off, please check tire pressure, oil quality and air filter
• advanced vehicle systems would be more specific
• my favorite, "Holly! Stop driving so fast, 1 gallon deducted!" (my wife is terrified of running out of gas)

Bob Wilson

 

Anything I can do to help Burnsy!

 

Nope. Fuel Economy: Why You're Not Getting the MPG You Expect - MSN Autos lists 44 mpg overall which was published before the MY 2008+ EPA method change. I also have a subscription to Consumer Reports' web site and the 35/50, 44 overall was obtained on an 04 Prius. On a side note, the 34/47, 42 overall on Touring was obtained on an 08 Touring.

They don't often retest cars for mileage if there are no significant changes. The will sometimes test another of the same car within the same generation but w/a different engine and/or transmission. Sometimes they do this in the same review (e.g. test a 4 cylinder and V6 version)

New EPA mileage figures discusses the MY 2008+ change and also references the above study.

 

Bob: I always appreciate your attention to detail and analysis. FWIW, I'd imagine you have these figures already, but just an FYI for others here, this is what CR got on an 01 NHW11: 30 city/49 highway, 41 overall. 48 mpg for their 150 mile trip.

I leave it to you to try to replicate their results w/your NHW11 from the limited info we have about CR's test.

 

I get between 63 and 65 mpg driving ~80% of the time on the freeway, albeit in rush hour which is like around town stop and go driving. I spoke with a Gen II owner last week and he said he averages ~48 pmg at best - clearly the Gen III has improved mpg.

I am curious why Gockam appears to start threads about the problems people are having with the Prius or a single article he can find that is trying to make a headline out of nothing - he should review the many articles and tests showing that Gen III has higher mpg and is the best hybrid available - maybe a non-Prius owner wishing they could be?

 

This exactly why I've had intense discussions with technicians, engineers, and math guru's about issues with the BT plate with GenII.

They can make it look like any way they want on paper.. .especially if the money is right,...."look at Enron and the government!"
Sometimes there are too many variables to explain the final results that reality proves.

When I first got my GenII based on my driving, I was only getting 44 - 46mph most of the time before it was broke in.. after broke in, it was easy to get 48 - 50... or even as high as 54mpg... I know people get more, but I don't go crazy... I do drive with traffic.. just mainly watch my accelerations and avoid braking.

Now with the GenII... its "never" below 50mpg and usually around 55 - 56 without trying and my highest so far has been 59mpg. I don't even have one tank of gas through it yet and I travel with traffic if not faster.

Once the Gen III is broke in, I"m estimating another 4 - 5mpg improvement.

When the dust settles, comparing vehicles with comparable miles, I'm estimating a 10mpg improvement at worst and 15mpg or better at best than the GenII.

Toyota was very modest to not throw a fit to receive a rating of only 50mpg.. it does that without even thinking about it.

I realize we are not in the dead of winter, but it's hot here in texas and the air is ran all the time except occasionally early in the morning.

 

only read CR... not any posts here... my sig tells you what i have experienced, but i need to qualify the numbers since more than 60% of the driving is not done by me on the SPM.

before SO started using car as daily commuter, i was getting around 53 mpg (this was also with 2004 tidelands) and my driving habits have changed little. in 2010 i am getting around 58. now, the big question will be what mileage hit i will experience this winter.

so, improvement could be as little as 3 mpg, but at the same time, the driving i am doing on the 2010 is mostly the city driving which is supposed to be where i am losing the mileage.

on the 4 major trips i have taken, my mileage has been very impressive. but i also only drive ECO mode only and i suspect that CR doesnt drives the car the same way they test Corvettes which is why their mileage sucked on the classics and will suck on any other car they drive

 

caught in traffic plenty and if at 20 mph, i would love that, even my SPM would thrive. but most of the time its like 5 mph.

at least, regen now runs at any speed. with the classics cutting off at 8 mph, it sucked. stop and go traffic KILLS your mileage because its all battery power to move and no regen to get any of it back PLUS conversion losses to recharge

 

Trying to get back on track - to explain why CR's City number for the new car might be lower than the old car:

Looking at it further, it looks like CR's 'City' test is actually repeated stop-start short trips with very long driver changes between each step.

"Our tests. These tests are stop-and-go city-driving simulations on our test track, which has a total of 18 stops and 4 minutes of total idle time. Top speed is 40 mph. Two different testers each drive three runs for a total of six 2-minute, 40-second trials on every test vehicle. Total test time is approximately 16 minutes."

- from the report cwerdna linked to above.

We'll never know whether the 2G test was done with the coolant thermos hot or cold at the outset, but I can easily believe that the coolant in the thermos was hot at the start of the 2nd to 6th runs. The Gen 3 car no longer has the coolant thermos but does have exhaust heat recirculation. We could be seeing the difference here between the two ways of managing engine warmup, which in this specific repeated short-trip scenario looks like it produces a worse result from EHR than from recycling hot coolant.

Over here, we didn't have the coolant thermos on the 2G but do have EHR on 3G. The changes to this have produced 28% better numbers on the EU fuel consumption test. The highway fuel economy improvement is only 10%.

 

44 mpg? seriously? I just got back from a road trip where I did EVERYTHING wrong for gas mileage. I had the windows open, the sunroof open (I wanted fresh air) but because it was 105 degrees out, I had the air conditioning on. I was going up and down mountainous roads, and driving at 75 mph on straight aways. Anyway, I can't imagine what else I could have done wrong. When the trip was all said and done my mileage was 47 mpg. When I use the car for my "normal" trips to and from work... I get around 49 to 50 mpg. Not sure what consumer reports is doing to get that low of gas mileage, but I haven't seen it that bad.. yet.

 

It would be fun to compare charts of the first month of ownership of the GenII with the GenIII.. especially if the climate and outside temp was near the same.