Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] Wind generator output versus wind speed.



As to your assumptions:

Both the source (fan) and the generator have fixed bladed rotors (fan blade styles). The load of the generator is fixed (100 ohms) and the output is fed to a voltage/current sensor and then to a program which calculates and displays the average power versus time. Once the orientations of fan and generator are optimized, that power is fairly constant. There is a separate device (rotational motion sensor) that is used to measure wind speed and we were careful to keep the distances and orientations as constant as possible.

I did not expect the generator power to wind speed to be cubic but was somewhat surprised when I found it (admittedly one time through--will redo the experiment this weekend) to be linear (including 0,0). My 'gut' was to expect something more quadratic (and indeed I've analyzed some data from a 2MW Illinois generator that fits nicely with a v^2 dependency), but realized we had a lot of transfers going on--ignored the electrical to fan, but wind to blade and blade rotation to generator--and also had a fairly narrow range of velocities (.35-.55 m/s). With John's more complete description it is clear that we have a very complex system. [Just last weekend drove by a new wind farm going up in Western Indiana with one of the towers very near the highway. I did note the twist in the blades and a bend at the end--sort of like a 747 wing. None of those turbines were yet running--near (and I mean NEAR) the town of Earl Park, Indiana.]

Rick


----- Original Message ----- From: "Brian Whatcott" <betwys1@sbcglobal.net>

This is a scientific observation, and I would like to understand it correctly.
Rick correctly relates wind speed and the available wind power as a cubic,
or perhaps even better, the difference of two cubics.

In his model environment, he has a wind source provided by a three speed
fan. One can suppose he measures the electrical power input to the fan
and also its output wind speed.
He may possibly see a cubic relation between electrical power input and
wind speed output on a three point graph, but probably not, because the
fan is (I imagine) a fixed blade arrangement. And he may be directing
the airflow into a generator fitted with fixed blades.
He may then be measuring the generator output voltage, and may possibly
be providing the generator with a variable load, which he can vary so as
to maximize the electrical power generated at given airspeeds?
I apologize for the number of assumptions I am putting into action here
- I have no insights into Pasco's recommended protocol as you may have
guessed.

It seems to me better to proceed from this example experiment,
and generalize from its data where possible.

Just reading this description of Rick's experimental arrangement
(in my imagined form at least) may already be allowing you to think,
"Did I really expect to deploy two energy conversions both nominally
involving cubics, but with numerous sources of speed sensitive
inefficiencies, and expect to see a neat cubic curve emerge?"

The same sort of considerations may also be applied to the
slide 11 curves - though one assumes that the windmill blades are
provided with variable pitch arranged for constant speed given
sufficient wind.



Brian Whatcott Altus OK Eureka!