Offset value of 0.35 (or 0.4) in windspeed readings

A typical anemometer transfer function is derived from linear regression analysis and uses the familiar function for a straight line:

y = mx + b

where:

m = slope
b = offset

The transfer function for the 40 anemometer is:

m/s = (0.765 x Hz) + 0.35

When the anemometer cups are spinning, a frequency signal (Hz) is output from the anemometer. If the cups stop spinning, the frequency goes to 0.

One can see that if Hz = 0, the windspeed value will register the offset value of 0.35 m/s.

Note that if the print precision in SDR is set to NN.N, the value of 0.35 will round up to 0.4 m/s.

This is completely normal and expected. There is almost no energy in the wind at low windspeeds, so this value will have very little effect on power projections.

Some have asked if using an offset value of 0 would be better or if the offset of 0.35 needs to be subtracted from each windspeed...the short answer to both questions is "no". This would not provide data as accurate as when using the published offset. In fact, if you change the offset, you must also change the slope of the transfer function. This is NOT recommended!

As always, one will want to look at 0.35 data readings carefully.

Is the value "real" due to calm wind conditions or has the anemometer iced over?

In the case of icing, you may want to filter the data from calculations - in the case of calm winds, you may want to use the data "as is".

At some point in your project, someone will want to produce a "power projection" based on the frequency distribution of windspeeds from your site and a wind turbine "power curve". Icing data would need to be excluded from this projection.