We are currently using 2 Honeywell Flame Detectors on our GE LM6000 to detect flame out. Is there a correct and proven procedure for testing their operation? Also what are the common problems which follow these types of flame detectors as we are starting to recieve nuisance detector failure alarms?
There is too much to be said in this limited space. Many of your observations and concerns with applications are appriciated. I am an ex-Honeywell engineer that designs and manufactures U.V. detector sensors and systems. I could be of help with the Flame Safe Guard (FSG) 129464n & m used in the C7035 and others.
Please visit my website @ www.specpointsystems.com
The following is a general comment regarding the use of flame detectors:
Starting with the three most prominent concerns, in my experience, when using (Loss of Flame) Detectors for flame safeguard systems:
a) If used in 2 of 2 logic, i.e., both must sense flame loss to cutoff fuel to burner. Hence, such an arrangement can lead to catastrophe if either one fails to detect a true flame loss condition. I investigated such a scheme involving 8 burners. 15 of the 16 detected flame loss detectors properly tripping 7 burners. One didn't. its burner stayed online. This resulted in a major catastrophe involving loss of lives.
b) If used in 1 of 2 logic, i.e., either one senses flame loss, then nuisance false-trip rate will double). I investigated a situation where
the MTBF (false-trip failure mode) was 11 days.
c) The "bathtub-curve" used to predict failure-rate is dead wrong. Such a curve will adequately show failure-rate associated with the detector's
electronics, but not the "mechanical" aspects, i.e., lens fouling, loss of scanner coolant, or sight-path degradation.
Of course, there are solutions to the problems noted above. For more detail refer to my paper "Probabilistic Risk Assessment of Equipment
Safety Systems." A copy may be obtained request.
A simple "thanks" is all that is required. Those of you who do not know the word ( I'm amazed at how many on this list are lacking common
courtesy) need not contact me.
Phil Corso, PE
(Boca Raton, FL)
I am presently working on R&M of control sytem of a Boiler for which the flame scanners has not worked for quite a long time. I think your paper will be of grat help for me. May I request a copy.
Anindya_Ray... As you had requested, the document was sent directly to your e-mail address! Please confirm its receipt!
Regards, Phil Corso
I have worked on GE MS5000/6000 turbines. I have built a Flame Detector Detector or Tester, it is a high impedance 300VDC supply with a micro
ammeter and a UV source. When the cell is connected (correct polarity is important, reverse polarity kills the cell) I can measure the cell current at different levels of UV. This is compared to known new cell values. We get two failure modes, some cells are high current without any or very little UV, others go blind to VU. This has saved us many flame detectors.
Colin would you mind emailing me in regards to the UV detector tester as I am interested in making one for our sites.
I would also appreciate a copy of thr circuit diagram and test procedure. And also if anyone can help, if the sensor is damaged how can I identify the details for a replacement bulb/diode. I have a scrap cable which has been
opened, the only number on the bulb is 76432.