When starting out in the hobby of building wind turbines, the question of what the wind is actually doing up there is pretty academic.
I started out thinking that way, but now that I have a windmill, and am actually putting the power to practical use, the question of where the best wind comes from takes on more importance. So after the fact, I've done a bit of research into the wind in my area. Though the specific results don't matter to anyone who doesn't live here, the method may be of interest to others, so I'm posting a bit about what I've learned and how I did it.
Starting off, I'll point out that MS Excel is all I really needed to crunch the numbers and plot graphs. On a 1.3GHz computer processor, I found that my first try at crunching data in the spreadsheet was very slow. My first try was very messy. I realized that I was trying to use Excel statistical functions to do too much, when I can instead use spare columns or separate worksheets to break the data down a bit at a time.
I live in Canada, where weather data is always collected in metric. Any US readers that need to converto to MPH can multiply the KPH by 0.62. All of the data I processed comes from an Environment Canada weather station, 25km (15 miles) south of me. There were 4 years of hourly data available on the website, free to download.
Strathmore AGDM Site Data - Environment Canada
The anemometer is on a 10 meter pole, which is actually shorter than my turbine's tower. I visited the site personally, and the area is flat and very clear of obstructions, so I would estimate its roughness is about RL=0.05.
My tower is close to a barn, and a shelter belt of trees around mye house. This plays havoc with my wind. I estimate the roughness at about RL=0.5 from the West and South, while it's relatively open from the North and East; RL=0.1 or so.
The first few graphs are common and relatively easy to explain. The histogram lists how often the wind blows at each range of speeds. Wind speed histograms typically follow a "Weibull" distribution, which means that there's a smooth bulge on the left side, trailing off to the right. There is a distinct "dent" in my site's Weibull distribution. Where this comes from will be more obvious after a few more graphs.
The next chart is the median and average wind speeds, broken down by month. The seasons really affect what the wind does.
This chart shows how often the wind blows from each direction. There are huge lobes from the north and south. Very rarely does the wind blow from the east. Scrolling through the numbers themselves, and from personal experience, the wind only blows from the east when it's changing from north to south or vice versa.
Going further with the question of the seasons, I broke the wind frequency down into groups of seasonal data. This plots four lines on the chart for comparison. Now we can see that the wind is much more likely to blow from the South in the winter and fall. In the summer and spring its direction is more variable. There is no season where the wind tends to blow from either the West or East, however.
The next thing to do with the wind rose is to plot the average speeds coming from any given direction. This makes a nice little "egg", but here the lobes don't quite match the lobes on the previous graph. The wind is more likely to blow from the SW, but when it's from the SE, it's stronger. Figure that. I think this is where the extra bumps in the Weibull distribution, the first graph, come from.
The last two charts really gave the computer lots of chewing to do. I combined the frequencies, average speeds, and directions to create a plot that gives weight to the winds that are both more frequent and blow strongest. Don't worry too much about the numbers on the chart - they don't mean anything anymore and I should have taken them off. It's what you get when you multiply 15kph by 7% - a meaningless number. The thing to take away from the chart is that the strong SW lobe of wind frequency does not line up with the strong SE lobe of wind speeds.
Any wind turbine site in my area must be clear from the North and South directions to properly receive the best winds. Obstacles to the West and East interfere less with the windmill and may be tolerated.
This last graph is a bit of a stretch of the data, but it's interesting to look at. I don't think I would site a wind turbine solely according to this kind of information, because it exaggerates what I did on the last graph. Again, the numbers on the axis don't mean much, the graph is only useful for comparison's sake.
Power that the windmill can collect is a function of windspeed cubed. So I cubed the windspeed in the weighting and charted it again. Now I see a huge spike of power available in the North, which dwarfs the power that the windmill will collect in any other direction. There are plenty of complicating factors in this, because a windmill cannot collect all of the energy in the wind. A few storms with strong North winds probably skew the graph beyond reason. I still want to post it, as "food for thought".
A possible extension of the number crunching could be to superimpose a specific windmill's performance curve on the average wind speed/direction/frequency data, and see what happens. It would flatten out the bulges, where the generator's power curve prevents the windmill from collecting most of the power.
So what does it all mean? to me, it means that the shelter belt of trees around my house to the West, and the barn to the South of my wind turbine, cut down slightly on the power it can collect, but the strongest and most frequent winds, which come from the North all times of the year, have a straight run at it and I get it all. Something that mitigates the blockage from the SW is that the trees lose their leaves in the winter, just when the SW winds are more common.
The location of my tower was picked for a lot of reasons, not just what I "thought" the wind was doing at the time. It turns out I couldn't have picked a better spot, really, without going MUCH farther away from the house and the barn, making the transmission of power very inconvenient.
I was most surprised to see that there is rarely much wind straight from the west at all. Around Calgary, we get "Chinook" winds every so often, which are strong, dry, warm winds dropping off the mountains to the west of us. I'm going to guess that the Coriolis effect takes hold by the time it gets to us, and may be partly responsible for the big SW lobe on the winter wind frequencies.
Another surprise is that this data does not quite line up with the Canada Wind Atlas data for my area.
This wind rose has no big lobe for the North wind. I do not know why. It does have a big lobe for the West wind, which is not visible in my local station data. Again, some explanation can be granted to the Coriolis effect. The data in the Atlas is for 30 meters above ground, while the local station data is taken at 10 meters. Things change, and turn counter-clockwise, down here in the weeds.
So don't believe what everyone tells you about what the wind does in your area.
Now that I've got the software running, I can delete the data and drop in data from a different weather station, and see what changes. I might get another big surprise!
Steven
