January 07, 2009
Method To Reduce Wind Farm Output Variation
Varying levels of wind farm energy output cause both short-term and long-term problems with demand matching. Some U Wisc-Milwaukee engineers propose a method to dampen the variation of output of wind turbines over shorter time periods.
Now, Asghar Abedini, Goran Mandic and Adel Nasiri at the Department of Electrical Engineering and Computer Science, Power Electronics and Motor Drives Laboratory, University of Wisconsin-Milwaukee, have devised a solution to the electricity grid susceptibility to changes in wind speed.
The researchers have devised a novel control method that can mitigate power fluctuations using the inertia of the wind turbine's rotor as an energy storage component. Simply put, they have created a braking control algorithm that adjusts the rotor speed so that when incoming wind power is greater than the average power, the rotor is allowed to speed up so that it can store the excess energy as kinetic energy rather than generating electricity. This energy is then released when the wind power falls below average.
This approach, the team explains, precludes the need for external energy storage facilities such as capacitors and the additional infrastructure and engineering they entail. Their method also captures wind energy more effectively and so improves the overall efficiency of wind farming potentially reducing the number of turbines required at any given site.
An obvious way to try to store wind energy is to make heavier rotors. This would give the rotors more momentum. But that would increase the mass needed for the towers that hold up the wind turbines and propellers. An alternative would be to run a belt of some sort from the propeller to the ground to have the heavier rotor and electric generator at ground level. But the belt to do this would cost more money. My guess is that wind farm design engineers have considered these ideas and I wonder what impracticalities prevent their use.
It sounds like this is a technique to reduce power fluctuations on a scale of seconds or a few minutes due to gustiness of the wind. It will not store energy for an hour or day long lulls. There are a lot of constraints on the allowable rotation speed of the turbine. First, they have to produce 60 Hz power synchronously to feed the US power grid, and there is a fixed gear stepup from the 30-60 rpm blades to the proper speed for the generator, 900 rpm probably. Second, you have to consider the blade tip speed and the noise it generates; a large turbine has has to keep its blade tips away from supersonic speeds. Even at slower rpms, the noise generated is comparable the whoosh of a vehicle moving a high speeds.
Storing excess kinetic energy in the rotors would be akin to using flywheels.
A number of companies build flywheels, in general they haven't made great progress despite the stronger materials and better design now available. And electric utilities which are testing battery storage ideas and designs aren't opting for flywheels either. So kinetic storage may not be all that dandy.
Vertical wind turbine designs would seem a better fit with trying to store wind energy; the heavier parts can be on the ground.
Unfortunately vertical wind turbines aren't being adopted. From time to time there are optimistic reports about their efficiency. But they would still have to be high to catch the stronger winds, perhaps that is the drawback.
Second part: I think any plan to store at the individual wind tower is doomed.
Conceptually a central storage farm for an entire wind farm seems much more practical for maintenance and reducing the total number of active components. Centralization does require a means for connecting the towers to the central storage. That means electrical cabling, you aren't going to do it mechanically.
Aside: the centralized power generation of solar thermal is why I like it.
In solar thermal simple mirrors solve the problem of focusing the suns energy at the generator. Gone is intermediate conversion to electricity of the wrong phase or frequency. No network of power cables. (Each mirror does need enough power and information to track the sun, and to be turned off when the central generator must go offline.)
So a spinning blade 100 feet in the air doing nothing is the answer, pure genius.
Popular Science recently had an article which showed an generator stack separated by clutches. As the turbine overspeeds, clusters of generators are linked, adding load. This obviates the need for an expensive mechanical gearbox, and allows the load to match input from the turbine. In other words, you have variable input as the wind gusts, etc. and more or less generator magnets are clutched in or out as conditions warrant.
It follows that you could use a similar arrangment for overspeeding the blades and storing energy mechanically as in a flywheel.
Personally, if I were designing a windmill, I would use a real flywheel storage system, where the rotor is optimized to store energy. Energy storage efficiency is a function of the mass at the rim of the rotor, as well as rotor speed. Storage energy goes up as an exponent of the rotation speed, hence very small fast rotors are efficient. In other words, a hybrid storage system, where excess electricty generated by windspeed is simply used to spin a flywheel storage system.
In a hybrid system such as I propose here, the flywheel rotor can be designed to its optimal roation speed, and the wind turbine can be designed to its optimal speeds. Staying within a design envelope is always the best engineering solution.