Generator theory

[bigheadted0]

Hi can anyone spend the time to explain to me the thory of how a generator controls its output power (both real and reactive). i understand that you can control reactive power export by altering the field exitation but do not really grasp how this happens. the generators i am interested in will be grid connected

 

[TurbineGen]

Output power is matter of torque. Speed is obviously determined by the grid (I'm assuming you are referring to a synchronous machine here). Thus once the generator reaches synchronous speed, any power added by the prime mover simply increases the torque. Since you know power is proportional to torque X speed, increasing power at the prime mover increases the output torque and thus generator output power.

Think of it like this: The power grid is a long train with many engines. Your generator is powering one wheel. You could increase the power to that wheel and the train speed wouldn't increase, but the output of the wheel would. There is simply too much inertia for one wheel of thousands to make much of a difference.

Now on to the reactive power. This is controlled by excitation of the rotor. A weak magnetic filed will produce a reduced voltage at the machine, and a strong field will produce a higher voltage at the machine. However you know that since it is connected to the power system that the voltage is (relatively) constant. Thus increasing the voltage of the machine above the voltage of the power system you are connected to (via over excitation) results in added VARs. Since you are on the generation side, the roles are actually reversed. You are now supplying VARS, not taking them. Thus over exciting the machine is supplying Lagging VARs.

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[electricpete]

The Jacobian matrix gives the relationship between P, Q flowing through an impedance and |V|, theta accross the impedance.

As the impedance tends toward inductive, the relationship becomes decoupled so that P depends only on theta and Q depends only on |V|.

Think of your generator model including syncrhronous reactance and possibly transformer reactance leading to some infinite bus... those impedances are primarily inductive so the decoupling assumption is good. Theta contros for P. Voltage controls Q. (or conversely angle differences correspond to real power flow and voltage difference corrrespdond to reactive pwer flower)


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[waross]

In simpler terms;
The field is highly inductive so the exciting current is highly reactive. When the a grid connected generator is under excited it draws reactive exciting current from the grid. When it is over excited it supplies reactive current to the grid. In both these cases the train wheel analogy is appropriate.
The power output is strictly dependent on the mechanical power input.
Now when the set is islanded, everything changes.
The excitation controls the voltage.
The load controls the power factor.
The governor controls the power in to equal losses plus power out.
If power in is increased beyond this point, the frequency will increase until the load/frequency curve and the prime overs load/torque curve intersect or something else bad happens.

Bill
--------------------
"Why not the best?"
Jimmy Carter