We are running 2 frame 6 GT (Only gas) DLN-1 in synchronization with grid (Floating). Complex load demand 48MW Both GT's are kept in Droop/ Preselect mode. We have 4% droop. We have provided high & low frequency grid power isolation. Also we had provide a logic that on isolation of grid one machine will come on ISO mode as per selection. If that machine trips second machine will come o ISO mode.

We had one unresolved experience. Once the grid frequency had increased due to external disturbance. In spite of our machines in preselect mode the load on both machines reduce by 8 mW. On high frequency (50.4 h) the grid de-synchronized and we went to island operation mode . But the machine did not come on ISO mode as per defined logic .On under frequency both machine tripped since both were on droop only.

When we studied, L90LL for load lower command is coming which in turn actuate L70LX5 high temperature spread lower which prevented L83SCDT droop to ISO permissive. So machine did not come on ISO.

Two matters is mystery for us

Why this (L90LL) prevailed since machine was on lower load. Also why in spite of preselect mode our machines could not achieve the preselect value is un explainable.

De-synchronization occurred in 4 seconds, with in this time machine could not achieve the preselect set value. Trip sequence occurred in 3 minutes. Within this time machine could not come on ISO. After tripping of one machine second machine also could not come on ISO. Can someone throw some light on these from their experience.

 

First, your configuration sounds relatively complex.

Second, operating a turbine in Preselected Load Control Mode on an unstable grid will not allow the turbine to respond as it should to grid frequency disturbances. In fact, it will respond exactly opposite of how it should when Preselect Load Control is active, and will actually contribute to the grid instability!

Third, and probably the BIGGEST glaring thing wrong with your scenario is that, trying to operate a DLN combustor-equipped machine in Isochronous mode is very difficult, virtually impossible, really. Especially if the unit isn't automatically switched to Lean-Lean or Extended Lean-Lean combustion mode when the transfer to Isochronous is enabled (which you didn't mention).

If you were operating the unit at 24 MW, it was likely in Premix Steady-State combustion mode. When the grid frequency went high, the CPD and the airflow through the machine increased. Both of these will not be "tolerated" well by the DLN control or combustion system, especially when in Premix Steady State.

You said that a high exhaust temperature spread alarm was keeping the unit from switching, but you didn't say if there was truly a high exhaust temperature spread. I'll bet there was because there was instability in the DLN combustors that resulted in high exhaust temperature spreads caused by both the change in air flow (which affects air/fuel ratio) and change in fuel valve position as the unit was trying to change load.

As you should know, there are some "deadbands" which occur when transitioning combustion modes during which time no load changes can occur (even though load is automatically changing!).

Operating a DLN combustor-equipped machine in Premix Steady State on an unstable grid is, well, necessary in some cases, but not very good for the combustion system. And can lead to some very unusual circumstances.

Throw in the attempted operation in Isochronous mode, and that's a very excellent recipe for lots of, "Why did <b>THAT</b> happen?" and, "I've never seen it do <b>THAT</b> before!" and, "Why did it trip?" questions.

Seriously, without a lot more information about the particulars of the control scheme in use at your site no one can probably answer your question with any real certainty--except to say what's already been said. And has been said before on control.com: Operating a unit in Preselected Load Control on an unstable grid is not going to produce the desired results. And, operating a DLN machine on an unstable grid is probably not going to produce the desired results. And, operating a DLN machine in Isochronous mode is a risky operation at best (risk of unstable frequency and turbine tripping).

 

Thank you Sir,

You had addressed many of the unanswered issues. Words speak of your knowledge and experience in the field.

The machine was running in Premixed steady state during the instance mentioned. The frequency increase has led to CPD variation - resulting in unstable combustion. As per the trip logger or alarm history from DCS there is no evidence of combustion mode change.

Regarding the Frame6 with DLN-1 combustor suitability in ISOchronous, in island mode it has performed the functions well even at load rejection due to industrial complex tripping.

Of course one machine was in ISO & other was in Droop. We considered the case of varying frequency of a disturbed grid also the system functioning has to be the same. Without Preselect mode & with synchronised operation with grid, the load of machine was changing sharply a variation in grid frequency. With both machine on Preselect mode we got better stability. But the above mentioned incident occured.

Now we are running the same sequence with reduced under-frequency and over-frequency limit (absolute value 49.7& 50.3 ) to desynchronise with grid. I would like to know, your evaluation of the system behavior in a situation where one machine is in preselect and other in droop running in synchronisation with unstable grid (Frame6 DLN-1). Generation side we are running one in Voltage control mode and the second in power factor control mode.

Thanks.

 

There are still too many things that are not known.

And, you have to do what you have to do to operate your machine stably.

When a unit is in Droop Speed control, if the grid frequency is unstable, then the load will be unstable. It's as simple as that. Full stop. That's the nature of Droop Speed control--and one of the intents, actually, of Droop Speed control.

When units are operating in Droop Speed control and the grid frequency increases, the units in Droop Speed control are <b>supposed</b> to lower their load. Because when the grid frequency increases it means there is excess torque being produced and to reduce torque is to reduce load which tends to decreases frequency.

When the grid frequency decreases for units operating in Droop Speed Control, the units in Droop Speed control are <b>supposed</b> to increase their load. Because when the grid frequency decreases it means there is a deficiency of torque being produced and to increase torque is to increase load which tends to increases frequency.

That's one major reason why units in parallel on an AC grid are operated in Droop Speed control (besides the fact they can't all be in Isochronous mode): to help maintain grid frequency by varying the load in the correct direction in response to frequency changes.

When you put your unit in Preselected Load Control to hold it's load stable, your unit is contributing to the grid instability because it should be changing its load to maintain grid stability and frequency. If it's not changing its load to try to help maintain grid frequency, and it's continuing to pump out the same load when it should be reducing its load then it's responding exactly opposite to how it should. It's actually making the problem worse.

Imagine if everyone did the same thing. The grid frequency would just keep increasing until more units kept being separated by their over-frequency relay. By increasing the setpoint of your over- and under-frequency relays, you are part of the problem instead of part of the solution.

Again, we just don't know enough about the situation at your site, except to say that if you feel that you need to keep your load stable during grid frequency excursions then you'll do what you have to do to keep your load stable. But, your unit speed is still going to fluctuate with the grid frequency, which is still going to cause havoc with the DLN combustion system, especially in Premix Steady State where the air/fuel ratio is very critical to unit operation.

I believe it's most customary to switch a unit operating in Premix Steady State to Lean-Lean when switching from Droop to Isochronous. That way, the fuel splits (and the air/fuel ratios) aren't trying to change quickly to respond to frequency changes.

But, it all really does depend on the particular circumstances of the site and how the plant supervisors believe the plant should be operated. And, still, we just don't have enough information to be of much help.

Without being able to examine a lot more data, and especially without understanding the particulars and the nuances of the site and the plant there's not much more to be said.

And wishing it were different (i.e., that more could be said) won't make it happen.

Best of luck with your investigation.

 

Thank you sir.