I've been watching an ongoing argument between several guys with very different backgrounds, on US grid frequency control (in the ERCOT area) and I think all sides are saying the same thing. It's their terminology/slang that's not consistent. I really don't want to get involved with the fight, so I'll ask the question here.

Using my version of terminology/slang....

Who controls the frequency on a large grid?

On a small grid, the unit that is big enough and with a governor that's on Isochronous Control (with Proportional/Integral Control) directly controls frequency (generator speed). Right? Others share MW load using drooping governors (Proportional Control using frequency (speed) as the PV and being fed a setpoint by a PI controller that is responding to its MW error...MW setpoint minus the MW's actually being generated).

On a large grid, no single unit is big enough to maintain frequency (no Isochronous Unit), so that job is shared among all units that participate in (get paid for) frequency control. The primary control method is through the droop characteristics of the "sharing" governors (Proportional Control...droop...on speed, and PI control on MW load as described above). And, the secondary control method is the "frequency obligation" (in MW's) that is added into the MW setpoint signal that is fed into the participating Automatic Generation Control (AGC) systems. Right?

In the Mark V, the droop is in the 4% droop control loop (P Control). Right? The MW loop is in the External Mode (PI Control) loop that gets its MW setpoint from the AGC system. Right?

Thanks,
Bill

 

Oh, and these "discussions" can get quite heated and downright ugly. One of this author's favorite sayings when describing these types of dialogs is "two people, separated by a common language." The use of slang terms (especially by those people from Texas) can be absolutely maddening--and one can be saying exactly the same thing, but they INSIST on using the down-home twangy, slangy description and forcing the other person to use it as well.

It would be nice if we could get some grid system operations personnel to comment on this, but it is believed you are correct. On a very large grid, also known as an "infinite grid," there most likely is/can be no one machine which is large enough to control grid frequency. It is believed that there are so many prime movers (all operating in Droop speed control mode!) that one 25 MW, or 80 MW, or 150 MW, or even a 600 MW machine can't have a serious effect on grid frequency (unless there is a serious deficiency of spinning reserve on the grid--and here comes the questions about spinning reserve!).

On a small grid, the Isochronous unit must be "large" enough to "absorb" the expected surges and drops in load (demand) to be able to effectively control grid frequency.

Grid system operators control the grid frequency, through remote dispatching using control schemes commonly referred to as AGC (Automatic Generation Control). AGC can take several forms--an analog signal or a digital (contact closures) signal. Grid system operators increase power output during high demand and decrease power output during low demand, to keep grid frequency "constant." (It does go up and down--just by thousandths or hundredths of a Hz, if the control is good.)

Speedtronic turbine controls (including Mark V) can have varying amounts of droop, but 4% is very common for most parts of the world. The droop is in the turbine speed reference vs. actual turbine speed portion of the control. The "outer MW loop" you are referring to is what's called 'Constant Settable Droop.' It is believed there is some reset (Integral component) to that portion of the droop equation.

markvguy