7E DLN1 Primary Re-Ignition Combustion Hardware Tempered on Cold Side

Has anyone ever seen this in the field? 7EA DLN1 with chronic PRI issues.
Fuel split at 83%, PRI at 1.5 hours. After first PRI, cycle to PRI gets shorter - 30 min..20 min…15 min. Combustion hardware had tempered like colors on cold side of Liner aft end and TP fwd end interface. At TP picture frame and on TP dilution holes on inner side. Around liner fwd end where mixing holes are. All cans look exactly the same. TBC looks like it got hot too. No cracks, spalling or bulging. I’m guessing because both hot side and cold side were seeing similar temperatures. N263 on TP fwd end above ring was a nice dark blue and the tractor seat lugs were a teal blue. Later stages do have fouling. Stage 1 had varnish stains making me suspect an oil leak from #1 bearing. Primaries have a thick coke deposit on center. We have replaced all combustion hardware (primaries, secondaries, liner, TP and all accessories. Still getting PRIs. Igniter depth is good, purge valves are good, no FOD in pigtails, fuel composition is good. Anyone ever experience this? See photo attached.


Also, Gas Only, No dynamic issues, NOx normal, no exhaust spread issues. CPD and CTP look normal. Exhaust temp is a little low compared to sister units. IBH behaving slightly different from sister unit. It opens sooner. On problem unit it will open when IGV is at 50 vs on sister unit it opens at 55 when all other parameters are the same. This chart shows IGV vs IBH. Red is the bad unit and blue is good. 1st pic is when both were running well, and second is when issues started. See photo attached.


ERG, in my experience of maintaining 2 DLN1 units for ~100K fired hours my issues with PRI events was due to low gas temperatures. We do not have a fuel heater and our gas temperatures were ~40 degf into the gas fuel module. After the pressure drop across the SRV and the primary valve, it was cold enough to cause frost on the flex lines between the gas fuel module and turbine package. The belief was that liquids were condensing out of the gas and providing a path for the flame to run back into the primary zone. You say your fuel composition is good, but not sure what that means. We were assured by our supplier that the fuel was "Dry".
Do you have a fuel heater, if so what is leaving temperature setpoint?
What is the inlet gas temperature to the gas valve valve enclosure?
Fuel split of 83% is slightly higher than optimal, but not out of spec.
Are the PRI events happening at any consistent load or ambient conditions?

Some other odd comments that don't mean a lot, but let me ramble.
If you had an oil leak from #1 bearing I would expect that to show up in the accessory compartment, there is no path into the compressor airflow path for that bearing that I am aware of. It is my understanding that if the telltale drain line for the #2 bearing fails then it is possible for oil to get into the CDC case and merge with compressor airflow. I have had a break in the telltale drain line of one of my units, but never had oil into the airflow path, just lots of noise and air out of the telltale line outside the package. If you are getting oil into the compressor airflow path that could explain your primary events.
Although I have never seen it, insertion depth of the spark plugs is called out as a possible cause of PRI events. I would assume you were able to measure and visually verify depth into the liner while the can cover was off?

I fought completely random PRI events for 5+ years, tuning, split checks, borescope inspections etc. that never turned up a thing. Since we have been able to maintain inlet gas temp >85 degf the issue has gone to bed. Hopefully I have not cursed myself by saying that. We were very close to purchasing the high load premix transfer software from GE years ago since this was happening so often as a way to keep from exceeding our emissions permit. Good luck and let us know if you have more info. A very difficult issue to diagnose for sure.