Most reliable pilot "flame on" sensor? 3

[molecularman]

We currently have two flares (30 inch and 12 inch) and are considering some options to try to improve the reliability of our pilot sensors.

Currently both flares use thermocouples on each pilot and of the five total pilots only two of the thermocouples are providing accurate readings.

We also have Infrared cameras looking at the flares in general but I don't know if it is acceptable per US EPA regulations to verify that there is some flame present (be it the main tip flame or any one of the pilots) at the top of the flare or if you have to verify that in fact the pilots themselves are burning. My reading of the regulations suggest you have to know the pilots are lit and I don't think we can do this with our IR cameras. That said, it is hard for me to believe that if we know we have a flame present that the pilots would't also get lit. Any suggestions on this would be appreciated.

My second question is this: Have people had good experiences with reliablity of flare pilot flame sensing using flame ionization sensors (and ignition systems)? Acoustic sensors on the flame front generator line? Thermocouples? If you've had good success with thermocouples, did you have them installed in protective wells? Have you had good success with retractable thermocouples that can be changed out "on the fly"?

We are considering flame ionization backed up by either TC's or acoustics but I spoke to a very experienced flare guy and he says we could get very good life (5-10 years) using heavy-duty retractable thermocouples in high alloy wells. I tend to believe this is the best solution since I worked with an oxidizer that ran at 1800 F and we got 1-2 years life with these in wells and with heavy duty shielding.

Also, does anyone have any words of wisdom related to ultrasonic flow measurement of flare gas streams? We are planning to put these on both flare stacks.

Last, I wondered if anyone had any personal experience with IR-type flare smoke monitors and controlling steam assist using the same.

So - questions ---

Best bet on flare pilot monitoring?
Ultrasonic flow meter words of wisdom?
Smoke monitoring and control?

Thanks much for any help you might provide

 

[jwysmooth]

I am not a flare expert but I play one on TV, but there is a contributor here that seems to give good advice on flares. Below I have cut and pasted his response to a similar question.
----------------------------
Thermocouples - most common method, long experience so most record of problems. Most problems come in the associated connections and wiring because it's not been designed for the high temperatures. Don't use regular wire in conduit in hot zones, make sure the tails are heat resisting, use mineral insulated stainless or inconel sheaths, use thermowells, all wiring connections in hot zones with SS MIMS. Shop made wiring nds (not site made), double protect everything from the flame lick.
Only measure the pilot flame, not everything else.

Accessibility poor so get it right for at least a 5 year span.
Pilots and thermocouples can be mounted together on a guide track. helps for in-service maintenance but is a pain to design (and get down ?) because of the potentially long runs and the need to handle piping and wiring as the pilot moves.
Thermocouples can be "poked" up a long tube. Tend to sag out of place over time and not read the flame, difficult to keep straight during fitting, may need special tools.

Relatively short stacks can be fitted with a canting device to get the pilots and the thermocouple down in one piece. tailor made solution

Have done all the above

Ionization - hate it. great idea in an enclosed environment but susceptible to wet, dirt, rust, shorts and uses High tension ignition.
Stay with High Energy ignition if you want direct lighting. Works in water and soot!!??
But needs activation with (a thermocouple?)

UV - stay away sun light activated

IR remote optical fairly common these days. Benefit is remote mount. Drawbacks are remote mount sees fog and mist before it sees the flame. Distance limited. may not see the pilot because something on the tip is in the way, may really see the main flame most of the time (if you care), can't see individual pilots without individual devices

Acoustic - only one manufacturer uses, My view .. gimmick..
The theory is there but I'm not convinced on the practice
relies on acoustic filters to "find" the tone emitted from a specific device. Distance and roughness attenuation in a pipe are concerns (site tuning ?)
-----------------------------------
The original author's logon name is "flareman"
Your question is a little like asking what is the best car, Ford or Chevy? Much depends on the application.
Cheers,

 

[Flareman]

Thanks to jwysmooth for kind words and for saving me a lot of typing. I haven't changed my mind since that exhange.

I do have a comment relating to the "retractable" solutions. There are a number of different approaches to this which address the chance that you may need to change out a thermocouple before you can shut down the flare for access. All tend to be a little tedious to accomplish so you need to assess the mechanical details of the solution offerred. One of these solutions uses a single really long thermocouple pushed up a guide tube, which would seem on the face of it to be a good solution. One potential drawback is that the weight of the thermocouple itself can cause the 300 ft(?) long x 1/4" (?) diameter rod to buckle slowly and "settle", over time into the guide tube, thus shortening slightly, which may pull the tip out of the preferred location in the pilot nozzle. I don't have an immediate solution to this problem other than looking at the relative clearance in the guide tube and the initial stiffness of the thermocouple element.

I certainly agree that using thermowells will improve longevity. This is a way of avoiding direct flame contact with the relatively thin sheath material, a consideration which should apply to the entire installation. Historically a great number of thermocouple fail in service because the local wiring cannot tolerate the exposure to high radiant loads. Cable selection and location need to part of the solution if connections close to the tip are envisioned.

Can't help on ultrasonic flow meters!

Smoke monitoring is trending towards thermally activated systems. Some are "blind", being mounted around the flare tip at the top platform, others are optical systems for remote or ground mounting.
The basis is that a smoky flame has a greater radiant output than a clean flame of the same size (all those little black bodies). Similarly the radiant load increase with a larger flame. In both of these cases, there is a need for more smoke suppressing medium so when the radiant load goes up increase the steam (?), and vice versa. This makes for a control system known as "scan and bias" rather than the PID method commonly used with a set point. Unfortunately most control engineers only know PID and are looking for a set point smoke|no smoke. There isn't one, - the system has to hunt constantly. If this is not properly understood the system never really works and the operator chooses a PID setting which works most of the time, and "tweaks" it now and then. This eventually finds a setting which may extinguish the flare at some flows (don't want THAT!!)

David

 

[molecularman]

thanks flareman and JWYsmooth for your responses. That was very helpful. I spoke with someone from one of the flare tip manufacturers with a lot of experience and he agreed with what you said. I'd say so much so that I almost wonder if you are him! I have the same concerns about the tall flare and the retractable element wanting to slide back down a bit. It wouldn't take much to back out of the flame.

Do you feel like the ionization detector/electronic ignition combination makes the ionization detector a more reasonable technology? Or, are you pretty much anti-ionization detector?

 

[Flareman]

Molecularman

I like the combination of flame ionization and ignition in a closed environment like a furnace. It makes a lot of sense for things like domestic boilers and heating units (and was probably developed for that market). However, the ignition system works using high tension (like an auto spark plug) and we all know how that works when it gets wet or dirty (not well!!)

In a flare the system has to stay in one place for 5 - 7 years (or more) and be ready for service the entire time. If it gets rusty, wet, sooted up, or droops over to touch the pipe or pilot nozzle it's pretty much dead.
I know that some people use it but I'm not a fan, on those technical grounds.

I prefer a high energy ignitor approach. You can get those babies to spark in a bucket of water or a block of ice.

David

 

[Chentronics]

To improve ignition reliability, you may want to consider the diagnostic high energy ignition system offered by Chentronics. As noted, these systems can spark in high moisture and sooty conditions. In addition, the Diagnostic offered by Chentronics senses the release of energy from the exciter. If the released energy is less than expected (usually to the spark plug wearing out and "skipping"), the exciter closes a set of contacts or flashes an indicator. -This wear indication can prevent problems...nice to know the spark plug is going to fail before it actually does.