Ask Your Question

Frequency deadband in GGOV1 model

asked 2019-09-30 06:21:41 -0500

KS gravatar image

updated 2019-10-02 10:04:04 -0500

perolofl gravatar image

Hi All, I am trying to put a 0.1Hz dead band in GGOV1 model. Now the deadband is to be stated in terms of pu speed. I calculated 0.1Hz dead-band from frequency (say 50Hz) and got 0.002 p.u and plug it in the model. With this value my model is responding to a frequency less than 0.1Hz, i.e. if I put 0.5Hz frequency change then it is calculating the MW change for 0.4Hz w.r.t droop. My answer should be the model should not be responding up to 0.1 Hz then it should respond to the MW change in line with droop. Please could you spot what am I missing here?

EDIT, inserted graphs:

Injected Frequency Ramp: image description

Unit Output:

image description

edit retag flag offensive close merge delete


I don’t understand. In what way is the model not responding as expected?

perolofl gravatar imageperolofl ( 2019-09-30 15:46:08 -0500 )edit

It is not responding to deadband. It's MW response is incorrect w.r.t droop.

KS gravatar imageKS ( 2019-10-01 02:34:34 -0500 )edit

For instance, with 0.002 p.u (0.1Hz) db set and -0.5Hz ramp F change, it is not supposed to react at the instant of ramp F change and wait until 0.1Hz F drop is exceeded. It is not doing so. The MW response is less than what it should actually be producing with 4% or 3% droop and turbine rating.

KS gravatar imageKS ( 2019-10-01 02:39:53 -0500 )edit

You need to supply more detailed information. Not only say that the response is not what you expect!

perolofl gravatar imageperolofl ( 2019-10-01 15:14:01 -0500 )edit

Do you have an email id if okay to share?

KS gravatar imageKS ( 2019-10-02 02:46:26 -0500 )edit

2 answers

Sort by » oldest newest most voted

answered 2019-10-02 10:02:56 -0500

perolofl gravatar image

The response of GGOV1 seems ok. With a dead band of 0.1 Hz the model will only sense a frequency deviation of 0.5 - 0.1 = 0.4 Hz. Hence, the droop response is about 80% of the response without dead band.

When the frequency is ramped back to 50 Hz again, the regulator stops the control when the frequency error is below 0.1 Hz, due to the dead band. The turbine output is therefore "stuck" at a level above the original power at start of simulation. That's why the red curve doesn't go back to 54 MW at the end of the simulation. That level correspond to the droop and 0.1 Hz frequency error.

edit flag offensive delete link more


Thanks a lot@perolofl. Much appreciated.

KS gravatar imageKS ( 2019-10-03 16:07:46 -0500 )edit

answered 2019-10-03 06:36:21 -0500

jfconroy gravatar image

If you're expecting that the deadband is +/- 0.1Hz, then test the frequency response of the governor with 0.09Hz and 0.011Hz deviations. If you see no response for the first simulation, and then a very small response for the second simulation, then you can be confident that you've "found" the deadband. Do the same for negative frequency deviations. In field tests of governors where you're not sure what the governor transfer function is (e.g. an old governor), you would do this type of test to try to identify if there is a deadband.

edit flag offensive delete link more


Thanks a lot@jfconroy.

KS gravatar imageKS ( 2019-10-03 16:07:15 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

[hide preview]

Question Tools

1 follower


Asked: 2019-09-30 06:21:41 -0500

Seen: 74 times

Last updated: Oct 03