Frame 9 Unstable in Premixed Mode


Thread Starter



Our company operate Frame 9E engine with DLN1. Fuel type gas only. Last time we have an issue with unstable work on mid load. Engine jump from Premixed to Lean-lean on about 60Mw (55% of full load) When we unload engine from 70 Mw to 60 Mw CO raise from 35 ppm to 258. Is it sign of blow out?

My investigation show that fuel split settings was changed (probable by GE last year during mapping) from 81-81-81-81 to 73-76-79-79 percent. No changes has done in gas fuel valves or IGV settings. IBH switched on. I cannot get any information from GE why they did these changes.

Now I think to return to initial setting of fuel split but I worry about acoustic. There is no and combustion dynamic monitoring system. Could you please explain what kind of risks I can get if I change these settings without CDMS?

When did this problem start?

How long since the last CI or HGPI?

How long since the last off-line compressor water wash?

Has the ambient temperature changed significantly in the recent weeks?

Has the gas fuel make-up changed recently?

Is the gas fuel supply pressure stable when this occurs?

Are you certain of the emissions monitor calibration?

At what combustion reference temperature is this "instability" occurring?

Are the IGVs and the IBH stable when this problem occurs?

What Process- and Diagnostic Alarms are annunciated when this "instability" occurs?

As to the risk of hot section damage from high/excessive combustion dynamics, there is always the possibility that tuning without a CDMS will result in issues--especially if the problem has resulted from one or more of the possibilities above (and those are just a few of the possible causes for "unstable" operation. It could be that high combustion dynamics are the cause of the problem and changing the fuel splits could make the problem worse.
Thank you for your answer!

> When did this problem start?

About month ago but it is not usual for this machine to work not in full load. No any work or maintenance was done at that time.

> How long since the last CI or HGPI?

Engine has 15K running hours now but CI was postponed till 18-24K. And we make borescope every 2K to monitor combustion components conditions.

> How long since the last off-line compressor water wash?

5 month

> Has the ambient temperature changed significantly in the recent weeks?


> Has the gas fuel make-up changed recently?


> Is the gas fuel supply pressure stable when this occurs?

Thanks for this question! I did not check it but I will do during my next visit to site.

> Are you certain of the emissions monitor calibration?

I did it we quite new Testo 350 gas analyzer but probe attachment point was far from engine. So I'm not certain of values but dynamic of CO changes still looking strange for me.

> At what combustion reference temperature is this "instability" occurring?


>Are the IGVs and the IBH stable when this problem occurs?

Sure there was some movement in both systems but I think it was not cause but reaction.

> What Process- and Diagnostic Alarms are annunciated when this "instability" occurs?

Only alarm I got was from spark plug when (I think) engine start to ignite first zone.

Bob Johnston

I've had a look at a standard 9E DLN1 control spec. and at 1980 F you are operating very near to the Lean-Lean to Premix transfer zone (around 1950-1960). you really shouldn't operate there. there is a chance that you are nearly trying to go back to Lean-Lean. Could you raise the load a little and see if you can get to at least 2000 F and see what happens then.

Let us know what happens then.


I suspect your site is one of many which operate using Pre-Selected Load Control, and that your operators are selecting a load which is "dangerously" close to the transfer point from Premix to Lean-Lean, and when the unit does go to Lean-Lean the load/TTRF is such that it wants to transfer back to Premix, and the (vicious) cycle continues.

The operators need to be trained NOT to select a load which results in a TTRF which is close to the transfer temperatures (you're going to have to determine what those are for the unit at your site). Perhaps Bob Johnston could provide the names of the Control Constants so you could look them up and post them near the HMI monitor for the operators to consult.

But, do write back to let us know how you fare!
Dear Gleb,

let me first say that my experience is with the 7EA version that is used here in the US 60hz market. But I think for comparison purposes I can at least comment on your 9E 50hz machine.

DLN1 machines are typically configured to guarantee 9ppm Nox and 25ppm CO within the 80-100% load range. That range can be expanded by use of Inlet Bleed Heat if the unit is equipped. As you reduce load to the minimum load range for premix you will typically see a point where CO concentration will spike to a very high number. At this point the compressor is providing more air than required and the mixture in the combustion zone is very lean, near a lean blowout condition.

As far as I understand these standard units configured for 9/25 Nox-Co do not experience issues with combustion dynamics due to the pilot flame being maintained at the tip of the secondary nozzle. The newer DLN1+ units do typically require a CDMS system to monitor combustion dynamics.

Also in my experience the adjustment of the "Fuel Split" which is biasing the amount of fuel split between the primary nozzles and secondary nozzles during premix mode, has little effect on CO emissions at minimum load. Typical emissions tuning for a DLN1 machine consists of finding the optimum fuel split that results in the lowest Nox. Some sites have a split array that moves this split through the load range, some just have a fixed split. Then the field engineer will record emissions readings and certain turbine data points usually at 5 mw increments from minimum load to maximum load. This data is then sent to a "Tuner" who enters this data into a "Cycle Deck" program that GE created using turbine models. This cycle deck program will generate control constants for part load and base load exhaust temperature references and lastly for IGV temperature control. These numbers are what typically keep the unit emissions at the guarantee level. These numbers usually require tuning at least twice a year depending on the ambient conditions, or whenever combustion hardware is changed during a combustion inspection or hot gas path interval.

It is not unusual for a DLN1 machine to reach a point at minimum load where CO emissions spike. That has been the limitation of the DLN1 system where at minimum load the compressor is supplying too much mass air flow. Inlet guide vanes and Inlet bleed heat work to reduce the mass flow and allow the unit to be run at a lower load and still stay in premix mode. But there still remains a point of no return where there is just too much air. That is the point where the unit typically has to enter lean lean combustion mode. It is also the point where lean blowouts can occur, or the flame can "Flashback" into the primary zone.

So I would suggest that you get GE involved and request the unit be tuned. Based on the data collected they should be able to determine if the CO emissions are a result of changes in ambient temperature, or possibly due to some change in the condition of your combustion hardware. But I would not personally be changing any of the tuning constants without knowledge of what effect this can have. I hope you find this information helpful.

First, I would agree and implied that changing Control Constants without GE's concurrence is not a good idea--not for DLN-I combustion systems.

Second, you had asked about whether high CO would be indicative of primary zone re-ignition while in Premix Combustion Mode. It could be, but there would be a Process Alarm indicating the unit was transitioning to Extended Lean-Lean Operation (and was firing the ignitors to assist with re-establishing flame in the primary combustion zone of all combustors). Another indication of primary zone re-ignition is an increase in exhaust temperature spreads, and the Speedtronic also (usually) transfers from Premix to Extended Lean-Lean when high spreads are detected during Premix operation (and the Extended Lean-Lean operation Process Alarm, and ignitor operating).

Third, it could be that some unknown trigger is causing the CO emissions to increase earlier during unloading, and that would indicate--as MIKEVI says--indicate the transfer temperature may have to be changed. BUT, that should all be overseen by GE and done with their input and assistance. It's all part and parcel of DLN tuning. The number of hours you indicated would also be of some concern to me, but since you're performing borescope inspections periodically that would hopefully indicate combustion hardware problems before they cause catastrophic failure, though we don't know how long since the last borescope inspection. Sometimes, failures of combustion hardware can occur quite suddenly. I'm NOT suggesting this has happened, but just as a possibility.

If this is the first time this has happened, and it hasn't happened since, I would tend to think something else was happening which wasn't detected at the time which caused the problem. Sometimes inputs to the TTRF calculation can cause erratic combustion reference temperatures, but they are usually accompanied by Diagnostic Alarms (but not always).

Please write back to let us know how you fare in resolving the problem!

Bob Johnston

The transfer to Premix is initiated by FXKTS1 which should be around 1970 Deg. F. there is a deadband of usually 50 Deg. F. Your system settings may be a little different but this is at least a starting point.