Weird MPG

[Spektyr]

I just completed a round trip from Wichita, Kansas to Dallas, Texas and noticed a hard-to-explain jump in MPG on the return trip.

Starting from a full tank I drove south on I35 to the Sunoco in Pauls Valley, Oklahoma (because it has 93 octane) and then from there on to the Buc-cee's in Denton. On the return trip I filled up at Buc-cees and stopped at the same Sunoco in Oklahoma.

Same route both ways. Drove down on the 1st, back today (the 3rd).

On the way down the first leg (219 miles) averaged 28mpg, and the second leg (114 miles) averaged 29mpg.

On the way back the first leg (114 mile second leg of the first trip) averaged 34mpg and the second leg (219 miles) averaged nearly 35mpg.

Variables:
The second leg of the trip down was driven 4mph faster than the first leg (which is what prompted me to keep a close eye on things on the way back - faster should be lower mpg). The return trip was driven at the same overall speed relative to the speed limit. I picked a number I was comfortable with and set the cruise control to the speed limit plus or minus that number. So for example, if my number was 1mph, I would drive 76 in a 75, 71 in a 70, etc.

Wind was blowing out of the south both days, but I haven't heavily investigated how hard it was blowing along the route. So I had a headwind going to Dallas and a tail wind coming back.

Temp started around 80F in Wichita and rose to around 100F when I arrived in Dallas. It was about 104F when I left Dallas and about 101F when I arrived in Wichita.

There was about 100 miles of light rain on the way down as I went through Oklahoma, but mpg was consistent before, during, and after the rainy stretches.

Car had about 2150 miles on the odometer at the start of the trip, which totals around 700 miles.

I don't drive particularly aggressively, and I didn't vary my driving style significantly between the two trips. By filling up at the same gas station in the middle of the journey I (more or less) eliminated fuel quality as a significant variable. (I only run 93 octane, but maybe someone put the wrong stuff in a gas station tank.)

The countryside does slope downward as you go south as an overall trend, but it's only a few hundred feet over the course of 350 miles, so really not significant and I got better mpg going "up" so definitely didn't matter.


My theories are:
Wind played enough of a role to change mpg. Not sure if that's even realistic.
Engine break-in. Even though I had a couple thousand miles on it already, perhaps those 300 miles southbound were the magical "end" of some wear-in that significantly improved efficiency? Seems unlikely.
Temps - my brain says it remembers reading once that cars tended to be more efficient when it wasn't crazy-hot outside, but I might have that backwards.


Anyone have ideas/theories as to why I'd see a 5-7mpg jump in a couple days?

[fredzy]

It’s wind. I once got like 38mpg heading East with a strong tailwind for ~300 mi through Ohio in my WRX, which usually is Just over 30. It’s basically 95% of what the car is fighting at high speed, greatly outweighs rolling resistance.

[Narwhal]

Wind makes a huge difference. Used to see a difference of about 5mpg going on direction vs another for a 300mi drive.

[Dadhawk]

I'm on team wind as well, depending on how hard it was blowing. The rule of thumb I've seen is you lose about 13% efficiency for every 10MPH of headwind. Of course, if what you have is a crosswind, the calculation gets a little complicated, and I've never attempted to do it for a car. (I've calculated crosswind component for an aircraft but that isn't what we are talking about here).

To get a precise calculation it would involve the drag coefficiency of the particular model car and a bunch of other variables. I know about it in theory but have no idea how to calculate it.

[Tcoat]

There is nothing like a good tailwind to improve mileage.

Ask any pilot or ship captain.

[fredzy]

Tailwind/headwind makes an especially big difference when you have a small engine

[Stonehorsw]

I looked on the altitude map just ou of curiosity and seems like the influence of elevation change is not main driver here.
+1 for wind

[Tcoat]

Originally Posted by fredzy

Tailwind/headwind makes an especially big difference when you have a small engine

[Spektyr]

Yeah, I think you're all right.

It didn't occur to me until I started thinking about wind specifically, but I didn't really see the hood flutter on the way back. Maybe a tiny bit here or there, but nothing that stood out.

On the way down there was definitely a fair amount of dancing going on from time to time - likely when the gusts picked up.

Wind speeds were around 10-15mph following on the return trip, out of the south or southwest depending on where along the map I looked, and the drive was pretty much straight north so there really wasn't anywhere I wasn't getting a boost. The road snakes around but not enough to be really fun, and not enough to really have any stretch of time where you're broadside to the wind.

If I'd thought about it hard enough I could've answered my own question: the reason you do speed records with two runs - one in each direction - is exactly for this reason. Otherwise it would rely too much on who picked the right track on the right day with the biggest tail wind.