Flame Sensor Intensity in GE Frame 5 Gas Turbines

We have two GE Frame 5 Gas Turbines used for power generation running on Natural Gas as fuel. One has Mark V and other has Mark VIe control system.
Flame Detectors installed on both machines are of Make Honeywell Part No LG1093AA26 installed in combustors 7 and 8.
Recently we had a brief shutdown for Air Filter replacement. Flame intensity before shutdown was as below:
28-FD-7 = 848
28-FD-8 = 583.

During shutdown when 28-FD-7 was checked it was found faulty and was replaced with new one.
Intensity after shutdown is
28-FD-7 = 250
28-FD-8 = 600.

I have two questions:
1. What may be the probable cause of failure of FD-7 as it was working fine before shutdown.
2. Why the intensity of the new sensor is on the lower side. How is the intensity related to performance / health of a flame sensor. Can we relate the intensity number to the health of the sensor or it is more related to the flame condition in the combustion chamber.

Well, I would ask the question: How was 28FD-7 tested? Because the data you provided showed it was working fine before it was "tested" and found faulty....

I'm going to say that the likely problem with the currently-installed 28FD-7 is: wiring issue(s). Most people are afraid--very afraid--to touch the flame detector wiring and so when it is disconnected for testing, and re-connected, it can be very poorly re-connected. And, when a new flame detector is installed to replace a "faulty" flame detector it is quite often poorly connected. Every time the wiring is disturbed to safely remove and connect or re-connect a flame detector there is great possibility for making a poor connection. Including causing unintentional grounds because of excessive handling of the wiring when breaking and remaking connections.

Many flame detector terminations are made using set screw wire nuts, or "just plain" wire nuts. And "just plain" wire nuts are often not rated for 335 VDC. And set screw wire nuts are quite often not properly tightened.

How long was the "new" sensor sitting on the shelf in stores (warehouse)? Was it previously installed and returned to stores (warehouse)? Was it properly stored (humidity and temperature controlled environment)?

Is the isolation gate valve fully open after the "new" sensor was installed?

Have you checked ALL the wiring terminations along the entire length of the circuit from the flame detector to the Mark*? Including the shield drain wire terminations? (New, inexperienced technicians have been known to improperly and unintentionally connect the drain wires.)

Please write back to let us know what you find.
1. We test flame scanners by connecting them to a spare Mark V/VIe control panel that we have. A flame is shown to its lens and its signal / Intensity is checked on the HMI. It was working fine when the turbine was running. After its removal during shutdown when it was checked on spare panel it was not detecting flame.
2. Connection are done using proper Terminal Block rated for 335VDC.
3. New sensor is brand new. However its shelf life is approximately 18 Months.
4. Isolation valve was not opened or closed during the maintenance.
5. Please shed some light on the connection of shield drain wire. Should it be connected on control panel side or not.
6. At what level of intensity does the sensor gives NO FLAME signal to machine.
7. There is a slight difference in Honeywell Model Numbers of old and new sensors.
OLD Sensor = LG1093AA26
New Sensor= LG1093AA46
Does the last two part numbers show cable length only or there is more to it? Can this be the reason of different flame intensity?

You do know that if the polarity of the 335 VDC is reversed to a Geiger-Mueller style (Honeywells are Geiger-Mueller style flame detectors) that if a high intensity light is applied to the lens that the flame detector will be ruined....

Shield drain wires are a huge subject which has been covered MANY times in the past on control.com. (Unfortunately, with this new forum software many of the old threads don't display properly....) It doesn't make a bit of difference where a shield drain wire is terminated--as long as it's only terminated at one end. By convention (this is NOT a standard (Remember: The great thing about standards is there are so many to choose from. :) ), shield drain wires are USUALLY terminated at the control system end (i.e., the control panel side (to use your terminology)).

As for the response to your Item 6, I don't know how to answer that question because I don't know how you measure the intensity of the light source you apply to the lens of the flame detector when testing it. Honeywell should have some kind of data sheet available on their website or included with the flame detector in the box it is shipped in to provide that kind of information.

As well for Item 7--you need to work with your supplier to determine what the last two characters of the Model number mean.