Optimum speed for best mileage is 65 MPH?

[MrWizard]

I've been reading a bunch of stuff about improving your gas mileage and
happened across an explanation of why slower is not necessarily better,
gas mileage-wise. You've probably seen the advice that you should be driving
the slowest speed, at which you are still in the highest gear, for best
fuel economy. Typically 45-55 MPH on the highway.

The contrarian theory is, within a gear, your car gets increasingly better gas
mileage the faster you go (until the aerodynamic drag and rolling resistance
factors start becoming dominant). This is because in the lower RPM ranges
(<2500 RPM) the torque is still steadily increasing - faster than the fuel
consumption - until either the wind resistance overtakes any efficiency
improvements, or until the torque curve flattens out. This point is very
likely not at the slowest speed in top gear, but more likely nearer the point
of maximum torque. Unless you are driving a very very un-aerodynamic vehicle.
Which of course makes me think of my Ford Explorer.

So I thought I'd do little gas mileage experiment with my Explorer and see what
kind of mileage I get at various speeds, while controlling every other variable,
as much as possible. I was surprised to see that I get the best gas mileage at
65 MPH, not 55 MPH. I didn't drive any slower than 55 for an extended period
since you get run over (or shot) around here if you're driving too slow.

Has anyone else seen similar results?

Averages over a 20 mile test range, after a 10 mile warmup.
Speed and MPG data logged continuously on my PC and averaged later in EXCEL.
26.59 MPG at 55 MPH
25.04 MPG at 60 MPH
27.97 MPG at 65 MPH <<<
27.50 MPG at 67 MPH
26.68 MPG at 70 MPH
25.20 MPG at 72 MPH
24.64 MPG at 75 MPH

My very ordinary Explorer: 1994, 4x2, Original Eng, Auto Trans, 3.73 rear end.
Mobil 1 Oil. Tires at 33 PSI. 87 octane fuel. No PCM mods.
Replaced: MAF/O2/EGR/PCV/Wheel bearings/Thermostat/Optima battery/Platinium plugs.
Recently cleaned: IAC/PFE/Seafoamed engine to death.
Changes: Removed roof rack crossbars, Airbox mod.
Lifetime gas mileage over 14 years: A very ordinary 17.3 MPG. Hence the desire to figure out, "where the heck
is all the gas going?" I am amazed at how absolutely awful my city mileage is, and how high it is on the highway.

 

[steventadams]

What are you worrying about mpg for? If you're getting those numbers in an Explorer you should be driving everywhere and back again...

 

[mattadams]

Seriously... those types of numbers aren't really seen in most SUV's .

 

[Tbars4]

...Please include in your first post some specs on your X...Such as auto/stick, gears, tires, tire size, air pressure, anything added to the vehicle, etc..This would make your statements easier to understand..

...Over the years I have done similar records of my trucks and I usually start from the day I buy them...These are not records I keep but I use them to better the performance of my vehicles while also trying to optomise fuel economy...

..I test over long drives and city driving also, but usually do it per tank of fuel used in gallon's...I have noticed that each of my vehicles, even if they were exactly the same, would have a different "comfortable" zone at highway speeds..

...Being that most of my vehicles have been pretty stock Ford trucks, the norm I have found was between 60-65 mph for best fuel economy..I have had a couple trucks that liked running at 70-74 mph better and that was when the speed limit was 55...At 55 they would use a bit more fuel...

 

[aldive]

I have a few questions about your test.

(1) how was speed determined?
(2) how was mileage calculated?
(3) was this with a stock PCM tune?
(4) what fuel was used?

I don't feel a 20 mile evaluation is valid. You should go at least 100 miles and preferable 200-400 miles for accurate readings.

 

[gijoecam]

Indeed, all vehicles will be different as fuel consumption is based on much more than engine size and vehicle size. Final drive ratio, transmission gearing, engine model year and design, etc, all have an effect. Heck, even whether or not the roof rails are in place could have a measureable effect.

As mentioned, each vehicle will have its own sweet spot where it will obtain its maximum fuel economy. I've personally found that 65mph in my Ex also yields the best results (23 mpg averaged over a full tank). However, 65mph in my F-150 actually causes a decrease in mileage over running at 75. For whatever reason, 75 seems to be the truck's sweet spot. At 75, it'll get 20+, while at 65 it'll barely get 16.

I don't think there's any kind of broad generalization that can be drawn from any one test of any one vehicle. A vehicle with any given setup of powertrain, drivetrain, and tire is only going to squeeze so many miles from a gallon of petrol... But comparing that vehicle to anything but an identical vehicle is like trying to compare apples to atom bombs... They're just too different, in my opinion.

-Joe

 

[Tbars4]

..Quote: But comparing that vehicle to anything but an identical vehicle is like trying to compare apples to atom bombs... They're just too different, in my opinion.

...I don't think anyone was comparing different models, if I'm reading your post correct...

...If this was due to my statement, my comparison's were based against same model and setup's that I had and even they would show a great variable...I was the only driver and mechanic and they had about 10k miles between them...

 

[gijoecam]

..Quote: But comparing that vehicle to anything but an identical vehicle is like trying to compare apples to atom bombs... They're just too different, in my opinion.

...I don't think anyone was comparing different models, if I'm reading your post correct...

...If this was due to my statement, my comparison's were based against same model and setup's that I had and even they would show a great variable...I was the only driver and mechanic and they had about 10k miles between them...

Sorry, I wasn't clear on my thoughts... let me try and clarify... It made sense in my head, but I can see how it made no sense in writing. I also think some of my thoughts were about MrWizard's post, and some were more about yours... I didn't make the distinction... sorry...

It's my opinion that a general statement that, "Optimum speed for best mileage is 65 MPH" is a little too general to be applied to all cars. To an Explorer, maybe. To his Explorer or one identically-equipped? Probably.

I would lean towards an even more general statement: "Any vehicle will have a 'peak efficiency'."

For some vehicles, it'll be at 43.2mph, for some it may be at 100+mph. For a brick like our Explorers, Al's may be at 67 mph, where mine might be 63.7 because it's a 2-door 4x4 with different gearing and a shorter wheelbase, and yours might be at a perfect 65... Who knows? One key to maximizing fuel consumtion on the freeway is finding that sweet spot and maintaining that speed, and I think most of us agree with that. It's no surprise to me that it's around 65. My own driving has shown me that for the last few months.

Does that make a little more sense this time?

-Joe

 

[Tbars4]

...lol, and I thought your statement was because I was not so clear on my statement.....

....As for my sweet spots for my 2 current vehicles, the Ranger likes 70-75 on a straight long drive...My X on the other hand, it's sweet spot is if it makes it through and over any obstacle in it's way...

..Seriously though, my X likes to run at 75mph but gets it's best milage of 16.5 hwy around 65 too..

 

[MrWizard]

How I Measured My Mileage

I added a better explanation of my test truck and how I measured everything.
Keep in mind the highway mileage numbers are for only the time that I was
actually going that precise speed. I averaged 3-4 minutes worth of consumption
at each speed when I wasn't accelerating or decelerating.

In real life, you start a cold engine, get a lousy 10MPG for the first few miles,
stop for a few lights, burn 3-5GPH accelerating up to speed, eventually get to the
freeway, drive the highway miles, accelerating for hills, passing, braking,
stop for lunch, start your cold engine, accelerate, etc... You then take the sum
total of all of this, and call it your highway mileage. Aldive makes a good point
that one needs to drive a long distance on the highway to try and swamp out some
of this non-highway mileage, if you're calculating hwy mileage this way.

What I'm doing is, making a change, measure the MPG, change back, measure MPG, make
a change, measure MPG. This way you eliminate as many other variables as possible.
This also makes it possible to accurately measure small improvements. For example,
I modified my air box to improve the airflow. You can, in real time, view the fuel
flow as you open or close the additional intake and see a 3% improvement. This
would be nearly impossible to confirm by comparing fillups.

-
How I measured the speed and mileage:
Custom Windows XP program I wrote in Tcl/Tk that measures the Vehicle Speed Sensor
frequency (MPH), measures the fuel injector pulse width, divided by the pulse period
(GPH), and calculates the MPG (MPH/GPH=MPG).
The calibration of the MPH reading was checked by comparing it to my TomTom GPS
MPH reading and timing mile markers. (My actual speedometer dial in my truck reads
about 5% high, so it was not used. For that matter, the GPS also gives an incorrect
reading while climbing a steep hill. It only shows your speed perpendicular to the
Earth's surface. Maybe the more expensive ones compensate for elevation?)

The GPH was checked by recording each tank's worth of data in real time and then
comparing to the gas station fillups. A few percent difference here. The engine RPM
cal was checked by recording the injector pulse period with an audio recording program
and manually measuring the pulse widths and periods. (My truck's tachometer also
reads about 6% low. Guess your car gauges aren't very accurate.)

Anyway, I think I'm well equipt to make relatively precise measurements of fuel
economy. One of the reasons for doing it this way, rather than making a change,
and driving long distances to determine it's efficacy - is to try and remove a
lot of the other uncontrollable variables (traffic, terrain, driving variations).

But back to the original question, do most people see better mileage at >55 MPH?
If there is a drive to lower the national speed limit to 55, and one of the
more un-aerodynamic vehicles gets better mileage at 65, seems like a bad idea....

 

[aldive]

I have a few questions about your test.

(3) was this with a stock PCM tune?
(4) what fuel was used?

Those were not answered.

 

[MrWizard]

questions 3 & 4

Questions 3 & 4 were answered in the edited version of the first post.
Buried in the description at the bottom.

I always use 87 octane. No point in using a higher grade. I didn't modify the PCM. The '94s have a
very ancient PCM, OBD-I, and no easy way of (externally) modifying the code or lookup tables.
Besides, I figured I should understand what the unmodified vehicle does, before making a bunch of changes.
That's when I noticed I'm getting better gas mileage at 65, contrary to all the media advice to
slow down to 55. Thought I better ask around as a sanity check.

Those were not answered.

 

[Charlie]

I get best mileage around 55/60ish mph, going up to 65/70 reduces it greatly. When traveling long distances and not in a hurry, I will run on highways if they are similar in distance to taking the freeway.

 

[aldive]

I get best mileage around 55/60ish mph, going up to 65/70 reduces it greatly. When traveling long distances and not in a hurry, I will run on highways if they are similar in distance to taking the freeway.

Same here.

Anything over 65 and I see a great drop.

 

[Russ in CT]

Very interesting thread. You even got better mileage at 70 than at 55 (and thats 'real' 70, not optimistic speedo 70).

I was always in the 'you should be driving the slowest speed, at which you are still in the highest gear, for best fuel economy' camp. But my eyes are opened now. Since its so difficult to gauge the effect of highway speed on fuel economy (since as you point out there are so many other variables if you just measure miles vs gallons at fill-up), this phenomenon never occurred to me.

So would this mean that, contrary to intuition, that if driving at a steady speed below highway speed (say somewhere around 40), you wouldn’t necessarily get the best fuel mileage in top gear? That there would be crossover points where using a lower gear, and hence higher rpm, could yield better fuel mileage?

In the initial test that was conducted, the gear stayed the same, and all other variables were altered (the engine rpm, hence the torque & vehicle speed, hence the resistance (friction, aerodynamics)), in an effort to find the point of diminishing return. Now we’re pondering what happens if you keep the vehicle speed constant (hence the resistance (friction, aerodynamics)), and vary the gearing (hence the rpm, hence the torque & engine efficiency).

My brain was able to accept the fact that, up to a point, driving faster can get better mileage. But its just not ready to embrace the idea that lower gear can get better mileage (at a given constant speed).

But thinking about it, I think I answered my own question. Even though the engine might be more efficient at the higher rpm (imposed by the lower gear), its also spinning faster. I would wager that the better efficiency at the higher rpm doesn’t come near to overcoming the increase in consumption (over time) due to the higher engine speed. The whole reason driving faster got better mileage was because even though the higher rpm actually used more GPH, more fuel over time, is the vehicle also traveled faster, covering so much more ground over that time, so it got better fuel mileage.

Just driving the same speed in a lower gear doesn’t help anything simply because you don’t get to take advantage of the car going faster when the engine is going faster. In fact, that may be where the whole (incorrect) idea of going in the highest gear at the lowest rpm came from. If at any given steady speed, using the highest gear results in the best fuel mileage, its easy to see how one might (incorrectly) assume that traveling the lowest speed in the highest gear would be best too.

 

[aldive]

I have tens of thousands of miles evaluating gas mileage in the Explorer and can state without any doubt that driving at 55-60 will yield significant gas mileage boost over driving at 70.

 

[Russ in CT]

So, how many times have you driven Mr. Wizards Explorer?

 

[MrWizard]

Hope we're not getting too far off the topic....

First off, I have no doubt whatsoever that if Aldive says he gets better gas mileage at 55 with his newer SOHC Explorer,
than at 65, then he does. You only have to read his 6 year long thread on mileage to know he's thorough.

Not that I'm some expert in this field, and I'm probably talking about stuff I don't know much about, but...

You're almost right Russ, you *usually* get the best fuel economy at the lowest RPM for a given speed.
There is almost a linear relationship between RPM and fuel flow, with no engine load. The complicating factor
is once you add the load, the fuel consumption is the sum of these two parts. Your fuel consumption depends on the
(torque vs. RPM part) + (more RPM, more fuel, part). How efficiently your engine converts gasoline into torque
(the torque vs. RPM part) varies nonlinearly with RPM. (It looks like a mountain at low throttle angles.
The peak of the mountain is what hypermilers call the "sweet spot", since you are producing the most power for the least fuel.)
One of the reasons Continuously Variable Transmissions are more economical is they can adjust their "gearing" to keep the
engine in the most optimal RPM range. In our case, as you point out, it's better to go faster when the engine is going faster.

My theory on why I get better mileage at 65, than 55, with my older OHV Explorer is, my "sweet spot" of most torque-per-RPM
is fairly low in the rev range. Which is typical of truck designs that trade off a lot of high-end power for more low-end torque.
So you have a very steep torque increase at low RPMs. Very common for truck engines, not so common for car engines.
I have noticed though, a lot of the newer truck (and car) engine designs emphasizing high-end torque, in order to get the
HP specs to be much higher. You'll notice the HP and torque numbers have gotten a lot higher, but at much higher RPMs.
You can also change your low-end torque by changing the size of your exhaust, changing your intake/exhaust valve overlap timing,
altering the runner lengths from your intake manifold plenum, resonant tuning, etc.
It's really difficult to state unconditionally, 55 MPH is the optimum speed for all cars, when you have all these factors at play.

And just to really complicate things, the torque vs. RPM curves vary as a function of % load and throttle angle.
Your engine is most efficient at high load, high throttle angle. So, to get best fuel economy you need to be using a high percentage
of the potential torque. Otherwise, if your throttle plate is barely open, you get a lot of turbulence in your intake airflow,
and you get lousy efficiency. At some point though the wind resistance is increasing faster (x^2) than the torque increase and
your fuel economy goes to hell.

The question is, is this commonly at 55MPH, or higher? I see my fuel economy peaking at 65MPH,
and going off the cliff beyond about 70MPH.

 

[Russ in CT]

I've read and respected (and endevoured to apply) much of the Aldive mileage thread. But the broad statement that 'the Explorer' gets better mileage at 55-60 doesn't account for the differences in where the 'sweet spot' is from one different generation of Explorer, different engine, possibly differently geared Explorer to the next.

 

[aldive]

I've read and respected (and endevoured to apply) much of the Aldive mileage thread. But the broad statement that 'the Explorer' gets better mileage at 55-60 doesn't account for the differences in where the 'sweet spot' is from one different generation of Explorer, different engine, possibly differently geared Explorer to the next.

How many Explorers have YOU tested the effect of spped on gas mileage?