flare gas metering

[causeandeffect]

hi all,
I am looking for a flow meter to measure gas flow rate in flare stack. The meter must be largely non-intrusive (being in flare pipe to avoid blockage)and has good accuracy (within 2%) and good turndown (100:1). I was told that ultrasonic meters are not reliable when flow velocity exceeds 80m/s due to increased surrounding noise and as you know gas velocities in flare lines can be as high as 0.7 mach(about 200m/s).
any ideas ?

 

[JLSeagull]

Often thermal mass flow meters used in flare applications. These are available with one element and multiple transmitters for those who want to see the very low rate and the blowdown rate.

 

[dcasto]

Think about what you asked for 100:1 turn down 2% accurracy, you can't have both. accuracy is typically at full span, so a 2% accuracy says at any flow +/- 2 units. at min flow you'll read 1 unit +/- units. And to top it off if you don't have an online GC analyzing for physical properties, you don't know what really went through the line.

That said, one or more retractible pitot tubes will give good accuracy and turndown.

 

[Flareman]

causeandeffect
I personally like both the thermal style and the ultrasonic style (big $$ difference though). Both can be top inserted through glanded valves and occupy little line space. Watch out for upstream and downstream distance requirements. Avoid anything which takes a sample or has an orifice (such as a pitot) as it will almost certainly plug up unless it works with an integral clean purge which prevents inflow of flare gas.

As noted above, you have to know what you want to read.
If you are looking for an accurate reading all across the possible range (for revenue purposes (?)), you could easily need 2 (or 3?) independent probes with either design.

Most times, meters are used for a control function based on the (smokeless ?) performance of the flare tip. In that case you can narrow down your operating range considerably and probably find a single device to cover the range you need. If it is a control function, you really don't care about the accuracy, more about the repeatability.

Oh! If you are looking for accuracy, don't forget to look into Mol wt. and temperature extraction and readout, and ask for reliable calibration curves for a number of your possible gas compositions. You would probably be amazed how many people are using scale readings for a single calibration (air ?) and applying them as though they mean something to gases with a significantly different MW or thermal characteristic, without appropriate adjustment.

David

 

[fribby]

Check out Panametrics (a division of GE) they have got a pretty good flare gas metering system, transit time flowmeter, available at the moment. I think the model number is CEM68. They should have some info on their website.

 

[JLSeagull]

Panametrics is under GESensing.com. FCI and Eldrige Products are among thermal flow meters.