Need A Gas Flow Direction Sensor

[sparkyyy]

Does anyone have a suggestion for sensor to determine gas flow, mainly direction? The gas temperature is over 1000 C and contains condensable tars while heading to a cracker. We have tried a pitot to differential pressure sensor, but the tars diffuse through the lines to the sensor, condense, and cause damage or plug. The gas velocity is 0-10 M s^-1 and pressure is never over 5 WC.
Mainly we need to know if the flow reverses and fire a safety.
Any help would be appreciated!

 

[roydm]

Wow, that's too hot for ultrasonic.

How about something like a crude venturi, same shape in each direction.
Two DP cells, one forward the other reverse, or maybe just a center zero scale.

If the line size is quite large you could fabricate the venturi shape from sheet metal and insert it inside the line. The 3 tapping points could be oversize in the top of line.

Another thought was some sort of paddle connected to a strain gauge.

Interesting challenge.

 

[DRWeig]

That's a tough problem, but maybe someone in the forum has solved it, wait and see. Here are my thoughts, but I have not applied them to your particular application.

Perhaps a bi-directional target meter, or a pair of target meters facing opposite directions? It seems that if you're interested only in direction (not accuracy), a target wouldn't mind having a bit of tar on its face. Might be worth discussing with a manufacturer.

I'm guessing that a conventional swinging vane would stick rather quickly if tars condense on it. There may be a way to shield the pivot with a thin polymer, though.

If there are consistent suspended particulates in the gas, there is the possibility of a bi-directional ultrasonic time transit meter. The particulates must always be there, though.

Good on ya,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[DRWeig]

Oops, I guess I was typing at the same time as roydm, and I didn't notice the 1000°C. Scratch the ultrasonic comment. Target still might work.

Interesting thought on the fixed paddle with strain gauge rather than a swinging one with switch...

Good on ya,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[danw2]

It sounds like the tar deposition was a problem for the DP cell, not the pitot tube. An approach for keeping the deposits away from the DP cell is to purge the impulse lines.

I've never actually done impulse line purging, but it is not unknown, and has been addressed in instrumention publications in the past.

The lower diagram shows impulse purging lines for a slurry line, obviously with liquid, so the taps and associated purge components are low on the main line. For gas, you'd want to 'flip' it so the taps and stuff are on the upper half of the main line.

 

[danw2]

What's intriguing is you've already got the flow element installed, a DP cell, all you need is to get what you've already got working consistently enough to detect reverse flow while keeping the DP cell from clogging

 

[sparkyyy]

Thanks for the help guys, much appreciated!
I think I need to approach this problem either up or downstream.

 

[agameterman]

We build Venturi meters for Power plants that require high temperatures as mentioned above. Typically the Venturi will have an accuracy of 1.5%