Tonneau Fuel Economy experiment

I'm a new Tacoma owner (2012 Double Cab, 2.7L, short bed). A big part of the reason I bought the Tacoma was the fuel economy for a double cab, compared to other manufacturer's crew cabs. It frustrated me that there was no manual transmission available for the Tacoma double cab, which would have been even more fuel efficient, but it was still the best fuel economy option. And with Toyota, you can't go wrong for all the other reasons, right?

In any case, when looking into way to improve fuel economy it doesn't take long to get into the Great Tailgate/Bed Cover/Tonneau Debate. Mythbusters says only mesh tailgates help. Other studies show Tonneau covers help. Just about everyone agrees taking off the tailgate doesn't help, nor does a traditional bed cover/camper shell. At the same time, there seems to be agreement that the particular model of truck will affect the outcome and there doesn't seem to be any wide-ranging test to look at how different ranges of models behave.

I was willing to drop the $400-$700 for a Tonneau, but I wasn't going to do it if it didn't show at least a 5% increase in fuel economy. Frustrated with the lack of consensus I decided to take matters into my own hands and do my own experiment.

First the result:

Control: without a Tonneau cover I drove a 115.6 mile loop at 70 MPH on cruise control and used 4.776 gallons, resulting in 24.20 mpg.

Experiment: with a plywood sheet, cut to size, over the bed, I drove the same 115.6 mile look at 70 MPH and used 4.902 gallons, resulting in 23.58 mpg.

So, on a 2012 Tacoma double cab with the small engine and the short bed (I make no claims for anything else), a Tonneau actually HURTS ones fuel economy.

Frankly, I was surprised... but the numbers don't lie.

For those who like details on experiments, I'll add a 2nd post to the thread with the details. I'll also do a 3rd post with questions and thoughts on where to go from here in my quest for added fuel economy (air filters, mesh tailgate, etc.).

 

you need to do the test again. Follow the scientific method.

your results must be able to be repeated consistently.

 

Hauling that sheet of 3/4" plywood may have been taxing on your 4 cyl , added load = less mpg

 

yup i would run it 3 times each and average. But 115 miles is a lot. i would cut it down and run like 20 miles and do that.

 

I'll bite, since I like looking into these mpg issues.

Your drop of 2.5% doesn't seem very significant. A 0.62 mpg change could result from any number of differnet factors.

How did you get your numbers? Scangauge or ultragauge? Did you use the same fuel pump when refueling? Did you top off or stop when it clicked off?

What isthe difference in weight of the plywood vs a tonneau?

Not putting you down, but scientific method and statistics demand that certain specific details are addressed.

FWIW my mpg changes by 2-3 between fillups. My routes dont change, but my driing habits do vary.

 

Did he specify 3/4"? Which would be heavy indeed.
Also OP same temp, wind etc?

 

I'll help you out on the air filters. They don't help mpg.

http://www.fueleconomy.gov/feg/pdfs/Air_Filter_Effects_02_26_2009.pdf

Try jacking your tires up to max sidewall.

 

Here's a study that found tonneau covers reduced coefficient of drag in all cases, with some much better than others.

Obviously this is not necessarily the results on a tacoma but it suggests that they should.

Wind tunnel tests are a lot more controlled than any driving loop.

http://www.trucktestdigest.com/PDFfiles/SEMATonneauStudy2007.pdf

 

I'm not sure much can be said about the study.

The trouble with numbers - people place statistical significance to them when they are enumerative at best. We in the Western Hemisphere are masters of placing statistical significance to things; the race to nowhere is littered with numbers. But it's a global problem.

Sample size 1: a common problem.
Sample size 3: a common problem.
Sample size 5: a common problem.
Finding the right sample size: uncommonly done.

That said, the qualitative studies and estimates are not without some validity. The thing to keep in mind is that aspects of the study or the whole study could be suspect. Understanding that should help one find balance.

As noted here, the variables in "the loop" are enumerable and unknowable. Using a scientific approach will help to minimize the effects of variability and help to generate better data and analytics. There are studies out there to support various bed covers, but it takes time to find them and then, it's often not clear how scientific the effort was. So in the end, I think it lends to a better guess at what to do than anything absolute.

Based on reading I've done, the tonneau should help. The degree of help with FE will vary from vehicle-to-vehicle and type of cover used (vinyl, hardcover, plywood). The differences may be more-or-less significant. Justification is there, especially over the longer term and total outlay. And cosmetically, it may be more appealing. It is for me.

Good luck!!

 

Here are the details of the experiment:

The course: I drove from one Chevron that is just off I-80 in Rocklin, CA (cross streets Granite Dr. and Sierra College Blvd.) to another Chevron just off I-80 in Vacaville, CA (cross streets Merchant St & Alamo Dr.). This is a flat (about 175 vertical feet delta), straight, 56 mile section of a major highway (3 to 6 lanes along the route each way) and the two stations being right on the freeway ensures that I'm measuring a very consistent, nearly entirely freeway driving course. I fueled up on each end, a total of 4 times, taking the plywood off the bed of the truck for the 2nd loop

The methods to ensure consistency:
-Headlights off the entire time (minor FYI, my pickup has the daytime running lights)
-A/C off the entire time
-No music the entire time
-Filled up at the same pump on both ends
-Used the same brand and octane (91) of gas I always put in my vehicles
-Warmed up the vehicle (over 30 miles of driving) to get to the start
-Used cruise control set to 70 mph. Speed set via GPS speed, not speedometer. (note, I had to come off cruise control for a few moments a total of 4 times, twice on legs 2 and 3 and twice on leg 4.)
-Used GPS logging to track route for end verification
-Measured gas after first auto-fill complete and topped off once and measured again.

Data from runs:
Leg 1 (Plywood on, from Rocklin to Vacaville):
Temp: Start 54, End 55
Car Trip Odometer: 57.7
GPS Trip Odometer: 56.66
Gas after auto-fill completes: 2.145 gallons
Gas after one top-off: 2.366 gallons
GPS Average speed: 69.2
Trip Time: 49:08

Leg 2 (Plywood on, from Vacaville to Rocklin):
Temp: Start 55, End 57
Car Trip Odometer: 57.9
GPS Trip Odometer: 56.86
Gas after auto-fill completes: 2.305 gallons
Gas after one top-off: 2.536 gallons
GPS Average speed: 67.5
Trip Time: 50:42

Leg 3 (Stock truck/control, from Rocklin to Vacaville):
Temp: Start 57, End 63
Car Trip Odometer: 57.6
GPS Trip Odometer: 56.67
Gas after auto-fill completes: 2.065 gallons
Gas after one top-off: 2.291 gallons
GPS Average speed: 68.2
Trip Time: 50:29

Leg 4 (Stock truck/control, from Vacaville to Rocklin):
Temp: Start 63, End 63
Car Trip Odometer: 57.9
GPS Trip Odometer: 56.62
Gas after auto-fill completes: 2.249 gallons
Gas after one top-off: 2.485 gallons
GPS Average speed: 67.0
Trip Time: 51:49

(sorry, I'd put that all in table format if I could figure out how)

Some notes:
* I was disappointed with the consistency in the speed. Although I don't think it significantly affects the outcome, the freeway was getting just heavy enough in the last run that it slowed it down a tad and generally it's hard to get the cruise control to be exact over this long of a run. However, looking at the GPX files and graphing the speed, it was pretty darned consistent. There may have been more effect on the overall run just based on the time it took to get on the freeway and pull into the gas station.
* The fuel economy was better going towards Vacaville than on the way back in both cases. There are a number of possible explanations. Vacaville is the slightly downhill side. Perhaps the wind was biased that way (although I'd have thought the opposite). Perhaps the pump in Vacaville was tuned slightly more generously. There are others possibilities too.
* Nevertheless, the fuel economy for the control was better than the experiment in both directions, so the differential doesn't seem to affect the outcome.
* I was concerned about the consistency of filling the tank, however, the consistency of how much the top-off added, about 0.22-24 gallons, gives me confidence the fills were VERY consistent.
* I'm not sure if the rising temperature affects things... my guess is not much.
* I'm more concerned the heavier traffic might have had a slight benefit to the control, as I was running a bit slower for just a few moments and the reduced drag due to the other cars. However, I don't think it is enough to offset the results.
* But again, the best the previous issue would do would be to bring them back to equal, with the Tonneau of no benefit.

So there you have it. Any concerns about my experiment and it's validity for this particular truck model/configuration?

 

Oh, one missed detail based on comments already coming in... 1/2" sheet of ply, which weighed about 40 lbs. Looking online cloth Tonneaus with their frame tend to weigh about 30 lbs and hard tonneaus 30-50, so this was in the ballpark.

To be clear, the plywood was still in the vehicle cab (it broke down into pieces) for the control run. So the weight was the same across the board.

 

really cool. just read through it all.

 

The problem with short runs is that without a fuel guage, the consistency of filling at the pump throws off the numbers. You need to get to a pretty long run before that factor shrinks in significance.

As to running it 3 times, yes, that would be better. Actually, I guess I ran it twice already, based on filling on both ends. As for doing more loops, gas isn't free. At the point where I could rule out all statistical issues, I'd have paid for the tonneau in gas.

 

I think you need to do the experiment with the plywood on and off five times each with full tanks of gas for each leg (i.e. ten full tanks of gas). Drop the high and low figures for each, and average the other three for each of the cover on and off.

Start with your gas tank empty right when the fuel light comes on, then fill up the exact same amount each time. A longer process will average out many of the variables that will cause minor differences in short runs.

 

At some point the test uses more gas than the lifetime savings.

 

My results with an undercover on a 2007 AC were similar. Whenever the cover was removed the mileage increased a bit, maybe 0.25-0.5 mpg at most. Not scientific but my fuel consumption records are accurate and the results were observed over 100k miles with the cover off maybe 20 times during that period. The weight of the cover has some negative mileage effect, Cd effect is unknown without a wind tunnel but empirical observation supports your results.

 

Just read through that. Thanks for the link. It looks like a very positive experiment until you get to the end:

"Aerodynamicists often say that a 10% improvement in aerodynamic performance is good for a 2-3% improvement in gas mileage, based on an EPA road speed of 48 mph. Therefore, an improvement in Cd of 5.73%
could account for a 1.8% improvement in gas mileage."

A less than 2% increase in fuel economy is not very compelling, when the cost is so high. It's also MUCH lower than what the Tonneau cover manufacturers claim (10-12%).

 

Damn , you guys are thorough , I was joking about the weight of the plywood

 

Why? He's going to use 10 tanks of gas eventually. He already has the cover built. Why not just keep track of the numbers during his daily driving and see what they say? The 10 tanks of gas is a statistical sampling of the available savings that could be (or won't be) achieved over the life of the truck. And the cost savings obtained during normal use is what we are looking for anyway, not in an ideal laboratory setting where you maximize fuel economy in every way possible. The longer test run ensures that the averages are solid, rather than just a tiny snap shot designed for an expected outcome.

The only other way would be a double blind test where you wouldn't know whether you have the cover on or not, or whether you are even being measured. But that's unrealistic if you have a view out the back of the truck.

I drive 100 miles a day, every day of the week. I'm sure that I could achieve a worthwhile savings in fuel economy over the 300,000+ miles of life in the next ten years that I expect to drive this truck.

 

Yeah I almost posted that quote too.

Except it isn't clear if that is only at 48mph? There should be increased benefit that faster you go in relative terms.