The DLN combustor is unstable during fast load


Thread Starter

Dang Quang Hung

Dear all,

I am facing a serious problem with GTG frame 5001PA.

When the GTG changes operation mode from Droop control (the GTG load is 20MW) to Iso. control at 14MW, we are faced with two problems:

The first problem is the turbine trip because of Low frequency trip in 17/04/2007. When I checked the trip data it is so that the flames inside combustion is unstable event. Some flames were loss, the exhaust temperature increased very high then the SFR is in temperature control after 18 seconds Turbine tripped because of low frequency.
The last trip in 10/08/2007 the turbine trip because of high exhaust temperature spread.

After I checked the historian data of GTG I found that after changing mode from Droop control to Iso. control mode about 3 seconds the exhaust temperature spread increased very quickly, and after 60 second delay Combustion trouble alarm appeared. Then the DLN I combustion mode change from premix to Extended lean - lean mode. After 9 seconds the turbine trip because of high exhaust temperature trip.

Could you please explain to me why the combustion is not stable after changing mode of GTG?
When the problem happens, can I change the combustion mode from Premix mode to Extended lean - lean mode by pressing extended lean - lean push button on the HMI of GTG in order to make the flame more stable?
Is the DLN I combuster not stable when the load changes quickly?
DLN combustor-equipped GE-design heavy duty gas turbine-generators don't make very good frequency control units. Period.

It's pretty common for units which have to run in Isochronous speed control mode when separated from the grid to be switched automatically to Extended Lean-Lean or Lean-Lean and to be prevented from operating in Premix Steady-state combustion mode while in Isochronous speed control mode.

DLN-I Premix steayd-state combustion is a very fine balance of flame stability, air-fuel mixture, and NOx and CO limits. Sudden speed changes such as can and usually do occur when running in Isochronous speed control mode can have very bad effects on the balance, so bad that the flame can be extinguished very easily.

You should have an OEM representative analyze your situation and make some recommendations and modifications. It is NOT good for a DLN unit (any unit, really) to be running for any period of time with high exhaust temperature spreads even if it eventually trips on high exhaust temperature spreads. On a Frame 5 unit, the turbine rotor is spinning somewhere around 5100 RPM--that means that 5100 times per second the turbine nozzles are being heated and cooled, heated and cooled, heated and cooled, as the buckets pass out of and into and out of and into and out of and into the cold area where the flame has been lost. Simply not good to be thermally cycling the turbine buckets that way.

One of the recommendations which will probably be made is to automatically switch the unit into Extended Lean-Lean combustion mode if it's running in Premix combustion mode when the separation event occurs. If you're fast enough, you can switch the unit yourself by clicking on the EXT LEAN-LEAN target and then confirming the action.

But, do something if this is a frequent occurrence; or you'll be spending a lot of money on turbine parts.
Thanks for your commendation.

The control system of our gas turbine is Mark VI controller. When I check the software and the trend before the unit tripped, I found that if after the line breaker trip, after 2 seconds the spreads will occur and then the turbine will try to recover spreads for 60 seconds before a Combustion trouble alarm will appear. The control system will require re-igniting the primary zone. If after 9 second high exhaust temperature spread still present then the turbine will trip because of high spread of exhaust temperature. So I think that if I change the delay time to detect the spread as soon as possible, when the spread occurs unit will re-ignite in the primary zone of combustion to switch combustor mode to Extended lean - lean.

I do not know if the unit is working on primix mode if I change the mode to Extended lean - lean by re-igniting the primary zone of DLN combustor. Are there any problems which will harm the Gas turbine combustor?

Best regards

Rahul P Sharma

The grid freq at our site is pretty unstable... Does it mean that DLN combustors will be a bad choice??? We are in negotiations for MARK II to MARK VI upgrade along with some upgrade of Turbine parts. In this package DLN upgrade is proposed for the existing Frame V MS5001P model turbines.

Though we dont have to switch to or operate our machines in Isochronous Mode, but is it not likely that in Droop speed control mode, frequent changes in Grid frequency would result in similar changes in speed and by your description it can result in unstable operation???

au revoir
The admonition to "do something" should not be misunderstood.

There's lots of sequencing that can be modified in the field without factory review and consent that would not adversely affect the operation of the unit. But DLN and combustion monitor sequencing should NEVER be modified without the packager's or GE's approval.


Under any circumstances.

Not no how.

Not no way.

Not ever.


The Recommendation was to have someone knowledgeable in GE-design heavy duty gas turbine-generator operation and control review and analyze what's happening at your site and what might be done to improve the current situation.

But, I repeat, do not make changes to DLN-related sequencing or the combustion monitor sequencing. Customers who have done so have invariably gone to GE afterwards complaining about parts life or stability or operational problems. And when it's learned that the sequencing has been modified without review and consent, well, that usually ends the "relationship" pretty much for a long time.

I have learned the hard way that these Internet posts usually don't contain all the relevant facts and information to a particular situation, not because anyone is trying to hide or conceal them, but because they don't seem to be pertinent to the person writing the post at the time. Only by working with a knowledgeable person who can ask the relevant questions and analyze the entire situation can you hope to get the best response and recommendation, particularly when it comes to DLN- and combustion monitor sequencing. One should never make modifications of any kind to any system based on information obtained for free from an "anonymous" source without carefully considering and understanding every possible ramification, especially when we're talking about DLN- and combustion monitor functions of a 25MW (25,000,000 Watt) combustion turbine.

Speaking of the rating of a heavy duty gas turbine (we were, weren't we?), the output power available to drive the generator (which is how gas turbines are rated) is only about one-third of the total power developed in the turbine section of the unit. That's because roughly two-thirds of the power developed by the turbine is consumed by the axial compressor to compress the air required to burn the fuel. So, when you stop to think about it: A 25MW gas turbine is really a 75MW gas turbine! Only 25 MW of that 75MW of power is available as torque which can be used to produce amperes with a synchronous generator.

Please, think twice before making any changes to the DLN- or combustion monitor sequencing.

And then, think about it one more time.

There are two clarifications to my recommendation: DLN-I and DLN2.0 combustor-equipped GE-design heavy duty gas turbines don't make good frequency control units. Those with DLN-2.0e, DLN-2.6, and DLN-2.6e combustors are much better suited to frequency control.

Second, 5100 RPM is 85 revolutions per second (5100 / 60), so turbine buckets would be cooled/heated 85 times per second during a high spread condition (or 5100 times per minute). Not 5100 times per second, but still, an undesirable amount of thermal cycling.

The answer to your question can be found in Item 4 of

You will have to have people troubleshooting and tuning the DLN-I combustion system all the time (it's not something one can do by themselves!). Long-term operation in Extended Lean-Lean mode is not an option, either.

By all means, go for the Mark VI.