Does the Wind Affect MPG?

Cars today are so powerful that we rarely notice winds as we hurtle along at highway speeds, but believe it or not, your vehicle uses more fuel driving into the wind versus driving through calm air. Need proof? I believe the data below can prove it.

A Little Overview

Previously, I shared my evidence that air temperature affects fuel economy, and my evidence for wind speed comes from the same source. I've recorded the miles per gallon (MPG) I achieve on my commute to work each morning, and combined that data with wind speed measurements recorded by weather services reporting on the internet. Bear with me as I try to show the correlation I think is evident in the data.

The Simplified Analysis

This first chart compares the wind speed reported by a weather station in my area with the MPG recorded by my car. As you can see there's no real correlation between fuel economy and wind speed. What's missing is the direction of the wind relative to the car.

Intuitively, we expect a headwind to lower fuel economy while a tailwind should increase it. Without a directional component to the data, it appears to be random, and one might conclude that fuel economy is independent of wind speed.

Fortunately, the wind speed data sources used for this analysis include the direction of the wind. Given the wind speed and direction, the resultant direction of my commute based on the start and end points, and the average speed of the vehicle, we can calculate what I call the "Average Relative Headwind Speed." In simple terms it's the average speed of the air striking the car as it traveled over the entire commute, and combines the effect of the wind speed with the speed of the car.

This plot shows a stronger trend than the first plot, indicating that as relative headwind speeds decrease, fuel economy increases. You might also notice that it's difficult to represent the trend with one simple curve (shown here as a solid red curve). I believe that's because fuel economy is dependent on other factors besides wind speed, or temperature for that matter. In a future post, I'll try to combine all these factors in a way that better represents the connection between multiple weather conditions and fuel economy.

More Details on Assumptions, Methods, and Data
As I mentioned earlier, this data was gathered while commuting to work each morning in my 2013 Volkswagen Passat TDI SE which has a diesel engine and manual transmission. The MPG readings were provided by a dashboard monitor that computes average MPG over the course of a trip. I don't know how VW programs that calculation, but presumably it's a function of distance traveled and fuel flow to the engine.

The data presented here was gathered between February 9 and July 2, 2015 so that the date range matches my previous post, Does Outside Air Temperature Affect MPG? In that post, I noted that the route was consistent each morning until April 2, when road construction added a detour that included a longer 45 mph zone each day until the end of the data period. I've continued gathering data since then, so more data is available and will likely be shared in the future. As traffic and road conditions allowed, I attempted to maintain the same speed on each section of road each day, and tried to make my accelerations and braking consistent between commutes. Because I commute before 6 AM, traffic rarely interfered with my ability to maintain a consistent pace.

The direction of my route is simply based on a straight line drawn between the start and end points of my commute. The angle of that line relative to true north was calculated, and compared to the average wind direction during the commute to determine the angle of incidence with which the wind struck the car. Wind speed and direction were gathered from weather stations reported on wunderground.com, and chosen for their proximity to my route and ability to reflect the conditions my car experiences.  Unfortunately, winds can change speed and direction over the course of a 40-minute drive, so without more sophisticated ways of recording and calculating based on those changes, my results will suffer some amount of error due to this variation in the wind. I'm sure that shows up as some of the scatter in the data, but I don't currently know how to estimate that effect.

The Average Relative Headwind Speed (ARHS) was calculated by representing the car's average speed and direction as one vector, and the wind's speed and direction as another vector. The vector resulting from adding the two is the ARHS. When both vectors are pointing the same direction, the magnitude of the resulting vector is simply the sum of the two. When my car is driving directly into the wind, the magnitude of the ARHS is the difference of the two. Any angular difference in the direction of my car and the wind requires some trigonometry to achieve the ARHS.

The polynomial trendline shown on the second chart above was calculated by Microsoft Excel, and I chose a second-order polynomial because I know the force of the wind on an object is directly related to the square of the speed of the wind. I think fuel economy is probably linearly related to that force, so approximating the relationship between fuel economy and relative wind speed as a second-order polynomial made sense to me. I've wondered about the effect of the outliers on this approximation, but I'll leave that analysis for another day.

Conclusion
The effect of wind speed on fuel economy is evident in the data. The scatter in the data is greater than I'd desire, but likely exists because (a) my analysis ignored a variable(s) that also has a strong effect on fuel economy, and/or (b) my assumptions and measurement methods are too general. I am pleasantly surprised that the data shows the basic shape of a second-order polynomial, even if the coefficient of determination is rather low.

The next time we visit the MPG topic, we'll look at the effect of water on the road, and possibly some conclusions based on multiple variables.

Related post:
Does Outside Air Temperature Affect MPG?

Corn in the backyard, 2016 -- weeks 2 & 3

May 8

Although we had a generous amount of rain leading up to May 8, we accumulated about 20% fewer growing degree days than normal for this time of year.  Nonetheless, the corn plants grew to 2.5 inches, maintaining a reasonable pace, albeit slower than most years.

 May 15

For the week prior to May 15, our rainfall was adequate, and growing degree days were very close to average, resulting in growth that put the tallest plants at about 5 inches. Historically, that's below average for the third week's measurement, but about average for the middle of May. No reason for concern yet, especially considering the temperatures have been so cool.

The May 15 photos also prove that our days have not all been cloudy this spring.

Introduction to Modern Planting Technology

Planting season is in full swing in Central Illinois as most of the corn is planted and growing, while soybeans and other crops are not far behind. I was fortunate to witness the planting in the field behind our house, and experienced my first ride in the tractor while planting so I could witness some of the fascinating technology that has made farming more efficient than ever.

Farmer Wagenbach chose this 16-row planter that appears to be the same unit we've seen in this field before. The basic concept of the planter hasn't changed much in the last few decades, but there are several ways planters are controlled today that wouldn't be possible without computers and global positioning systems (GPS).

Here's a closer view of one of the planting units for those that are curious about the hardware at the ground.  While there are some features visible here that improve the seed bed in several ways, the technology that really caught my eye was in the tractor cab.

The bottom screen in this stack of three is a Precision Planting SeedSense monitor that tracks the performance of each planting unit and warns of detected problems.

The middle screen is a display connected to the auto-steer feature of the tractor. Most of the time I rode in the cab with Farmer Wagenbach he wasn't touching the steering wheel as the tractor was steering itself. After one picks a point at each end of the field on the first pass along one side, the computer draws a straight line between those points and establishes additional parallel lines all the way across the field spaced as wide as the planter being used. When the tractor is steered close to following one of those lines, the automatic steering feature takes over and keeps the tractor on that imaginary line until the driver takes control to turn the tractor around at the end of the field. In the end, all the rows are straight and evenly spaced.

The monitor on the top shows what portions of the field have already been planted and records all kinds of performance data for the planter that can indicate problems while planting, or can be compared to harvest data months later to study the effects of planting performance on yields. It's also an easy way to recognize slivers of land that aren't planted yet, so those areas won't be missed. Incorporated in the planter monitors is the ability to control planting units independently. This minimizes overlaps that result in reduction of yield.

Here's a spot in our field where the curved edge of the field on the right produced a small area between that pass and the straight rows on the left. This sliver was visible on the monitor so Farmer Wagenbach was able to pass over the area with the planter allowing the GPS-controlled planting units to drop seeds where the computer suspected none had been dropped before. The result is the five partial rows in the middle that, while not perfect, still drastically reduced the amount of overlap that would have occurred without this technology.

This caused me to be observant of planting patterns in other fields in our area, and it appears that several area farmers use similar technology. In this image, you'll notice the inside rows end nicely next to the outer rows that curve around them.

This is another part of the same field that shows the lack of overlap as the rows in the middle of the field on the right intersect with the rows that follow the edge of the field on the left.

For contrast, here's a neighboring field that was planted the old fashioned way with overlap in these tricky areas along the field edges.

Finally, another example that I found impressive. This one had the inner rows running diagonal to this edge of the field, which would have resulted in lots of overlap in the past, but looks almost effortless as planted this year.

Corn in the backyard, 2016 -- week 1

May 1

The ground had warmed enough by the middle of April this year that many of the farmers in our area took advantage of the dry weather and planted their corn.  Farmer Wagenbach was included in that wave, and I not only witnessed his planting in the field behind our house, but I had a chance to chat with him a bit and ride along in the tractor on April 19 (more details on that in a future post).

I first noticed the plants emerging on April 29, so our first Sunday corn photo of the year was taken two days later. The plants were easier to see with the naked eye than they appear in the top photo, and the bottom photo illustrates why: they're only an inch tall so far!

After the seed went into the ground, we received a little less than twice as many growing degree days compared to an average year, so even though it has been cool since the plants have emerged, they had a warm bed in which to germinate. We'll try to keep a close eye on the crop again this year, and I''m sure we'll be amazed again at the wonder of plant growth.