Recently, we had observed over frequency on Frame-V GTG having GE make Mark-V speedtronic control. Machine was paralleled with Grid power for about 15 minutes and later made independent from Grid. Frequency of GTG got increased to 52.1 HZ and resulted in tripping of synchronous motor loads.

Can Reactive load throw off on Turbine (GTG) would have resulted in speed increase?

 

Manoj Baviskar,

To answer your direct question: No; it's not possible for reactive load throw-off ("rejection") to cause an overspeed.

It's more likely what happened was that the turbine was operating in Droop speed control mode when it was synchronized to the grid, and then when it was isolated ("islanded") the Mark V did not--or was not--switched to Isochronous speed control, and so the frequency increased because the load was much less than the amount of fuel that was flowing into the turbine's combustors.

Without some means of controlling fuel to tightly control speed (which is essentially what Isochronous speed control is) a generator-set will almost always exhibit the kind of behaviour you witnessed when isolated from a grid--regardless of the governor (control system). Droop speed control is NOT tight speed control--in fact it's very name, "Droop," means that the actual speed is allowed to "droop" or vary from the speed reference. Machines operating in Droop speed control rely on other machines to control frequency, and when operating on their own in Droop speed control mode do a pretty bad job of automatically controlling frequency.

Hope this helps!

Now, I have a question: Does this happen frequently, the turbine-generator being isolated from the grid and having to supply an "island" load independent of other generators and their prime movers? Because if it does, what most site usually do is have a contact from the tie breaker (the breaker that connects the turbine-generator and its load to the grid with other generators and loads) to automatically switch the Mark V from Droop speed control mode to Isochronous speed control mode--and then there would be no problem with frequency control.

 

Many Thanks CSA for the prompt reponse,

Coming to your question:

The turbine-generator (GTG) is operated in "island" with Droop control and supply load independent of other generators and their prime movers. Also, Grid is taken as back up before large HT motor is started or GTG load is about to exceed Base Load. After above condition is normalised, Grid is taken out. Well, above operation happens quite frequently, but we are experiencing over frequency problem for the first time.

Please find the additional details wrt incident,

- Frequency raised from 50.07 Hz to 50.42 Hz in 7.5 minutes period

- Frequency raised from 50.42 Hz to 52.15 Hz in 3 minutes period

Now, my questions are,

1) Why high amount of fuel was flowing into Turbine combustors when the Active load (MW)was matched to GTG supplying an "island" load before isolated from Grid power?

2) Why it took @ 10 minutes to raise the frequency?

3) Is it recommended to switch over "Isochronous" mode when GTG is operating in Island?

4) If the GTG is running in Island mode, at load close to its base load in Isochronous control, what would happen when the load on the machine increases? Will the machine continue to run at the same frequency without accepting any further load or will the machine trips at some point?

 

Manoj,

1) The fuel was not high at the time of isolation--it drifted higher. There are two possible reasons: First, the Active Load decreased over the time period in question. And, second, the fuel valve is developing a sticking problem--could be the valve or dirty oil causing a problem with the servo-valve.

2) See answer to Question #1

3) I find it all too common for people to operate their turbine in Droop mode when isolated from the grid. If that's the way your site wants to operate, that's how it will be operated. I find that trying to change the minds of operators and their supervisors (who were usually operators before they became managers) is virtually impossible--actually it is just impossible.

4) When the exhaust temperature reaches the limit if the load continues to increase the frequency will start to decrease. And when that happens the frequency will start to decrease even faster. When the under frequency trip setpoint is reached (usually a generator protective relay monitors this and initiates a turbine trip) the unit will trip.

And the same thing would happen if the unit were operating in Isoch mode if load exceeds Base Load (exhaust temp control).

 

Your reply is an eye opener. We will have to seriously start thinking of Isochronous control for our Island operation

Please clarify following wrt your trailing comment

> 1) The fuel was not high at the time of isolation--it drifted higher. There are two possible reasons: First, the Active
> Load decreased over the time period in question. And, second, the fuel valve is developing a sticking problem--could be
> the valve or dirty oil causing a problem with the servo-valve.

DATA :
Droop setting for our GTG is 4%.
Overfrequency setting for downstream synchronous motors is 51.6 Hz.

Frequency raised from 50.07 Hz to 51.6 Hz i.e. 1.53 Hz = 3.06%.
This 3.06% change in frequency with GTG droop setting of 4% will require huge throw-off of active load. Practically, we had observed was load was dropped to 13.5 MW from 18 MW and it was the loading of synchronous motors prior to tripping.

This was the reason why we were doubting on Reactive power throw-off. Please provide your valuable guidance on the issue.

For info.: GTG had undergone Major Inspection @3 months ago and Lube oil was replaced with new along with it's filters.

 

Manoj,

You seem to be providing conflicting information, and you also say, "...Practically, we had <i>observed</i>...." You said in an earlier post the frequency had increased to 52.15 Hz, and now you said some tripping had occurred at 51.6 Hz, and the unit has 4% Droop (I imagine that you read somewhere that the gas turbine has 4% Droop, but have you verified 4% Droop?)

You said, "...the Active load (MW)was matched to GTG supplying an "island" load before isolated from Grid power...." I read that to say the GTG load was matched to the island isolated load before the grid tie breaker was opened. That would have meant that the frequency of the GTG when the breaker opened and was isolated from the grid should have been approximately 50.0 Hz (50.07?). Hence my statement: The unit was (apparently) operating at rated frequency when the isolated from the grid. If the operator did nothing over the time (10.5 minutes) and load decreased (which you say it did--from 18 MW to 13.5 MW) then the frequency would have increased.

A drop of 4.5 MW (from 18 MW to 13.5 MW) on a unit rated for 20 MW with 4% Droop would equate to <b>approximately</b> 25% change in load, which would mean <b>approximately</b> 1% change in speed (we don't know exactly what your machine is rated for, nor what condition the machine is in (axial compressor cleanliness; inlet filter cleanliness; exhaust duct backpressure; etc.) so these numbers are <b>approximate</b>. That would mean the speed would increase roughly from 50.0 Hz to approximately 50.50 Hz.

Observations won't do it; you need to be citing from recorded data--actionable data as it's called. Real values recorded by the Speedtronic or the DCS or the Historian or some data archival system. Observations--and eyes and lighting and oscillations--can cause huge discrepancies. Which may explain the numbers you are reporting--rather the differences in the numbers.

If the unit has 4% Droop, the frequency should not go above 104.0% (52.0 Hz)--and that's for relatively low load (not zero load, but some load). And you reported 52.15 Hz; how is that possible?

And what trips at 51.6 Hz? The turbine? The synchronous motors?

If you're operating a single unit isolated from the grid supplying an isolated load, then typically they are operated in Isoch speed control--so that the turbine control system will change fuel automatically in order to very tightly control frequency. To synchronize back to the grid, one needs to switch back to Droop control to synchronize, and then operate in Droop mode when synchronized to a grid. When the grid tie breaker is opened, the unit should be switched to Isoch control--that's how a plant should be operated. That's the purpose of Isoch control: to tightly control speed as load changes to maintain frequency <b>when isolated from a grid.</b>

But, the other stuff--with the apparently conflicting data and "observations"--it's very difficult to say for sure. Extremely difficult.