Archives

Here's some data showing how time of day affects my fuel economy on my commute. Even a few minutes can make a measurable difference. My commute is about 5 miles. For most of that, I deal with moderate traffic on a suburban highway, three lanes each way. Heavier traffic results in using the brakes a bit more as well as having an increased chance of hitting red lights triggered by traffic on intersecting streets.

http://i255.photobucket.com/albums/hh137/fanamingo/Driving/CommuteToWork_preEBH.jpg

http://i255.photobucket.com/albums/hh137/fanamingo/Driving/CommuteHome_preEBH.jpg

The trend lines above are just linear best fits. If I leave a half an hour later in the morning, I'm likely to lose about 2 mpg. Half an hour late in the afternoon drops my fuel economy by about the same amount. Sure 2 mpg isn't a huge hit, but it's about 5% of my average. These trends reflect small but noticeable increases in overall traffic volume. It's something worth paying attention to for those of us always trying to improve our numbers.

Now all I have to do is stop hitting snooze in the mornings.

I love charts like this. Not only are they informative, but they give me something to show other people when they think I'm obsessed with my charts!

Thanks for sharing this interesting data.

Ha! That's one I haven't charted. I think I have just about everything else, though. ;)

Great charts. Perhaps you could also plot average speed for the commutes.

Then make another chart with average speed on the bottom and mpg on the vertical.

I'd love to see the data behind those charts. Looking at them I am inclined to say the in the first one there is no trend, and in the second one the trend is questionable. There is just too much data variation to get those measurements.

If you did those in excl give us the trend line coefficient and the standard error on that number. That is one way to see if the trends are real.

I feel the charts do give good information even though the variation is large. It may be thrown off by the two points way to the left.

It might be interesting to log average speed, temperature, time you left and mpg for each trip. Also the day of the week as traffic around here varies a lot based on the day.

Even if it isn't accurate, it is fun. I might do something like that myself.

I'd love to see the data behind those charts. Looking at them I am inclined to say the in the first one there is no trend, and in the second one the trend is questionable. There is just too much data variation to get those measurements.

If you did those in excl give us the trend line coefficient and the standard error on that number. That is one way to see if the trends are real.
You're right. I put the data into charts and got excited that they looked pretty, but didn't check the statistical significance of the trends.

|slope|standard error|R^2
chart 1|-91.8|74.6|.03
chart 2|-76.7|32.4|.15

It's been a while since I've done statistics like this, but it seems to confirm you suspicions. There's not much if any real trend in the first chart, and the second seems to show a trend, but is not very precise. Here's the data if you really want to check. ;)

Chart 1 Chart 2
Time Mileage Time Mileage
06:36 40.9 15:07 36.5
06:38 40.9 15:20 39.0
06:44 40.1 15:45 34.2
06:44 39.7 15:45 38.2
06:45 38.5 15:53 32.7
06:46 39.7 15:53 35.7
06:48 39.7 15:53 33.1
06:50 38.2 15:53 36.3
06:50 41.0 15:54 39.0
06:50 37.7 15:54 30.3
06:50 42.1 15:54 35.2
06:50 38.2 15:55 34.3
06:51 38.7 15:57 32.9
06:52 40.5 15:57 33.1
06:52 37.5 15:58 34.5
06:52 38.2 15:59 36.0
06:52 37.8 16:00 35.2
06:52 31.0 16:00 38.4
06:54 38.4 16:01 35.4
06:54 34.6 16:02 35.2
06:54 36.0 16:02 37.3
06:54 33.8 16:04 32.5
06:54 35.7 16:08 32.6
06:54 40.1 16:09 34.2
06:54 38.7 16:10 36.8
06:55 35.1 16:11 38.4
06:56 41.8 16:12 35.2
06:56 36.3 16:15 35.2
06:57 36.0 16:17 31.6
06:57 36.2 16:20 31.8
06:58 41.1 16:22 35.6
06:58 42.8 16:23 34.1
06:59 37.8 16:25 32.8
06:59 34.1 16:25 31.8
06:59 36.9
06:59 39.4
07:00 36.3
07:01 37.6
07:01 34.4
07:01 42.2
07:01 38.4
07:02 35.5
07:03 39.1
07:03 38.8
07:05 39.1
07:06 40.9
07:07 41.1
07:07 33.7
07:09 39.1



It might be interesting to log average speed, temperature, time you left and mpg for each trip. Also the day of the week as traffic around here varies a lot based on the day.
Unfortunately, I didn't record average speed (or trip length). It would be an interesting comparison though. I also didn't record temperature, but for the morning at least, they should all have been between about 65 and 75 F. Evenings would have varied a bit more. I don't think I had enough data to get meaningful results from looking at each day of the week individually. A couple days only had 4 samples each.