The golden rule of power system operation

Say P - active power
     Q - reactive power.
     f  - frequency
     V - line voltage.

Then the golden rule is,
P is directly related with f.
Q is directly related with V.

Even certain experienced staff think that parameters such as line voltage depends on the active power produced and frequency may drop if there is not enough reactive power. This is Totally WRONG.

In other words,
If you are going to change the frequency of the supply you must increase the amount of fuel (eg. steam/diesel/HFO/waterflow) which in turn increases the prime-mover (eg. diesel engine/ gas turbine/ steam turbine/ hydro turbine) speed. But speed governors are meant to regulate this type of frequency variation and to maintain the speed of the prime mover.

If you are going to change the terminal voltage of the supply you must increase the excitation given to the alternator.

But keep in mind that the above are purely applicable to generators running isolated/islanded. Parallel operation and infinite grid operation are bit different and certain parameters can not be independently controlled (for eg. voltage in infinite grid). 

If multiple generators are running in parallel, only by increasing the excitation of all generators - the voltage can be increased, and vice versa. If not, only the reactive power share will change, not the output voltage.

The other important aspect of operating a power system is to do with the power factor. People confuse with whether the power factor must lead or lag. There is another belief, improving the power factor means we try to make the 0.9 to 0.8 lagging. In almost every practical power system (there are few exceptions) , the power factor should be lagging, BUT not necessarily each and every generating set. In other words, even when the system power factor is lagging - one machine may be running with a leading power factor, at the expense of another.