ROCOF settings

[PeterS84]

Hi All,

Can somebody give me any guidelines on what settings to use for rate of change of frequency? Would you normally do some simulations before choosing settings or just use a 'generic' recommended one.
Also, would you tend to use normal ROCOF or frequency supervised one?

 

[waross]

What is your application???
Is this for a large scale load shedding scheme?
I have seen simple underfrequency relays used for load shedding at a large plant as a condition of supply from the utility.
A drop in frequency was taken as an indication that the grid was overloaded and the plant was warned to curtail usage at 58 Hz. If the frequency dropped to 56 Hz there was an automatic trip of the main breaker.
Are you using rate of change to discriminate between events that may be ridden through and events that demand an immediate trip?

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

 

[PeterS84]

I'm thinking about using it as a faster alternative to frequency based load shedding. In case of small inertia of the system load shedding based on frequency may come too late to make any difference (as it takes time for the frequency to fall below setpoint). With ROCOF I was hoping to gain some time.

 

[Marmite]

While it is possible to estimate the rate of change of frequency from knowledge of the generator set inertia
and MVA rating, this is not an accurate method for setting a ROCOF relay because the rotational inertia of
the complete network being fed by the generation is required. If you have a system with a small inertia then the frequency response will be faster. You should be able to come up with some underfrequency settings that dump the load before the system collapses because the frequency excursion will be rapid with a small system. If you try to implement ROCOF on a small system you will be very susceptible to false trips. It is more typically used as anti-islanding protection rather than for load shedding.
Regards
Marmite

 

[PeterS84]

Thanks for the answer.
Why do you think there would be a lot of false trips? If I set the relay to, let's say, 5Hz/s this is some serious ROCOF now and shouldn't happen under healthy conditions. Unless what the relay is measuring and what is actually happening on the system doesn't align...

 

[PHovnanian]

You might want to do some simulations (analysis) to see how close faults will affect the voltage vector shift. You'll need to set a time delay appropriately to differentiate between actual system acceleration/deceleration and sudden phase shifts which present themselves as high instantaneous df/dt.