Flame rods in coker furnace 1

[Zinskie]

We have had issues with intermittent signal from 9 of 28 flame rods installed in our coker furnace. Through discussions with FireEye, the suggested solution was to remove extraneous cabling and rerouting to avoid any crosstalk on the Belden coaxial cable. Wiring modifications have been completed, but the problems persist.

While the furnace is down, we have been using an acetylene hand torch to check the flame rods' operation with decent success, save the last 5 elements in a row, furthest from the sensing relays (approx. 150 feet of cabling).

As it is an entire bank of fire sensing elements that are not returning a positive signal, I do not believe element faults to be the cause.

When testing with the torch, the flame is in contact with no fewer than two good electrically grounded components and the element covered with flame from 1/2" to 2".

Any suggestions on identifying potential faults would be appreciated.

 

[danw2]

Presumably the re-wiring isolates the flame signal wiring from the ignition wiring.

I assume the installation worked OK for a period of time, but has only recently given intermittent drop-outs? Or is it a new installation in commissioning phase?

What DC current reading do you get when doing the torch test? Normal DC on the working burners is ? ? ?

150 feet is the number Honeywell uses for the limit of flame rod wiring distance, before the 0.02uF capacitance (flame signal to conduit ground) limit kicks in. I'm not sure what Fireye recommends, but if you've talked to them, they must have taken that into account. The use of coax should minimize the capacitance in the run.

Are you using the shield of the coax to provide a wired ground between the burner ground electrode and the flame detection electronics ground? Or does the installation depend on device frame ground for the ground side connection? If so, has something changed to effect that ground continuity? A test would be to run 150 feet of copper wire and hard wire a ground between the electronics and the burners.

Your test with a torch rules out lazy flame or insufficient flame contact on the rod. But, are you sure the torch flame hit a sufficient ground electrode area? Rule of thumb is that the ground electrode area needs to be 4x the flame rod electrode flame contact area. Is the ground electrode area still intact? It hasn't burned or corroded away? Sometimes it's a collection of short rods welded specifically to provide a ground area.

Other possible common mode failures:
- AC supply voltage from a common source? Too low? Open circuit? Wire insulation cracked from heat exposure over time?
- Combustion air fault creating rich flame sooting conditions. (is the flame rod OR the insulator sooted? one is high DC resistance, the other an AC leakage path)
- ground electrode surface area has burned away or is too corroded or sooted to conduct
- electrical noise from VFD produces sufficient AC to trip the flame sensor electronics. Were the combustion air blowers recently converted to VFD control for fuel air ratio control ?

A long shot:
- flame rods replacement 'repairs' made with low carbon steel look-alikes: coat hanger wire, brake tubing or welding rod, all which corrode really well in combustion service.

As always, what's changed? Recent rebuild? Repair?