Flare Gas Measurement devices

[airbus]

What have been your experience with Flare gas flow meters? Our refinery (2 Plants) has three flares and measures the gas going into the flares stack (GE Sensing) after Water Seal Drums. On the case of one of the Plants, on ocassions of events of high pressures, when both flare gas meters should have read flow, one does and the other doesn't. The one that does, it reads two to threefold the amount being released (per our estimates). What type of technology do you use? Have you considered other technologies? Thanks, Airbus

 

[786392]

Just a wild thought,use ultra sonic meters!

Best Regards
Qalander(Chem)

 

[JLSeagull]

The GE Sensing meter is ultrasonic. I have one on order for a flare application now. A previous project used a two-range FCI thermal mass flow meter. It worked well the last that I heard.

 

[camabee]

Depending on the range of gas quality this can be a standard application for thermal mass flow. For large flare lines you may need multi-point sensor to get the accuracy you need. Try FCI, Kurz, ABB etc.

 

[VikingUK]

Ultra sonic meters are useless at low pressures (<5 barg) and low flow rates. If you want to measure accurately I suggest you fit a pitot tube.

 

[zdas04]

Pitot tubes are occasionally an option, but each one must be "calibrated" against a known standard in situ. That is really hard in a flare-gas situation.

I've also had very poor results from Ultra sonic meters at low pressures. Especially in a line to a flare where you just don't know what the make-up of the flow is going to be from moment to moment.

I would insert a plate in the flow (about 0.6 beta) and put in pipe taps to sense dP and upstream pressure. It doesn't have to be AGA-3 measurement, and with decent secondary devices you can easily get better than 5% uncertainty. You are going to be hard pressed to beat that in the real world with gee-whiz gizmos.

David

 

[786392]

Somehow I feel that any make-up gases may easily be metered at source their injection point pipeline probably with standard orifice or ultrasonic instruments.

Whereas actual off-gases from plant shall be checked/measured through multiple basic measurement element like pitot tubes,ultrasonic meter Calibrated& verified in-situ for varied load conditions.

Hope this helps better in way forward.

Best Regards
Qalander(Chem)

 

[JLSeagull]

I hope that VikingUK is wrong.

Like an orifice plate, the pitot tube is just another head type instrument. You would require multiple dp transmitters to accommodate the turndown.

As previously stated, thermal mass flow measurements such as FCI with multiple transmitters for range turndown is another technology used. However this project went ultrasonic.

My ultrasonic application is specified to comply with the rules relating to volatile organic compounds (VOCs) in chapter 115 of the Texas Commission on Environmental Quality. I hear that they are not very understanding when things don't work.

The fabricated ultrasonic meter specified for my flow application is to be flow tested in a third party lab at 30%, 60% and 90% full scale with an accuracy requirement 5% flow. Full scale is around 250,000 lb/h. Multiple outputs and Ethernet data may accommodate the low end, perhaps less than 250 lb/h. The transmitter accommodates six 4-20 mA outputs. I anticipate that the Ethernet data would accommodate the turndown better than six analog outputs. We don't plan to bring six mA signals to the DCS. You have to specify everything including the maximum pressure, temperature, range of molecular weights, ... etc. Include transmitters to provide process pressure and temperature to the ultrasonic flowmeter transmitter. My application is specified below three barg in a line sized 20 NPS and with a maximum dp specified below 10 inches of water. The probes are retractible. Our inspectors will test the flow meter per our testing criteria.

I am betting that the ultrasonic flow meter will work. However, I my specification data sheet lacks a specified requirement to work.
;-)

 

[zdas04]

I hope you're ok, but I've not been a fan of ultrasonic meters in varying pressure/varying gas content applications at low pressure. The changes seem to be to frequent and too subtle to be picked up.

Depending on the pipe size and how far your meter is from the burner (i.e., how much dP after the meter for max flow?) you can get some very large pressure swings as a cascading failure scenario causes several PSV's to open into the line simultaneously. I try to take this max flow into consideration when sizing flare header piping, but I see a lot of documentation where the engineer was really optimistic and/or just didn't do the arithmetic.

David

 

[dcasto]

Good luck. What I see is a 1000 to 1 turndown and that won't happen, digital or not. If your gas has the same density as air, you'll still be exceeding the maximum velocity in the 20" meter that most ultrasonic meters claim.

 

[JLSeagull]

As a flare application the molecular weight can vary widely. I show from as low as 0.02 lb/ft3 up to 29 lb/ft3. I think that the testing will use air. All of this is on the data sheet. GE Sensors is really pushing this technology for flare applications. The technology selection was made by others but I prepared the requisition. Instead of numbers like 1000 turndown our requirement is 5% flow at 30% full scale.

 

[VikingUK]

I actually have a turbine meter in my LP flare line. I would recommend it but for one problem, if you have condensed liquid in the line the blades stick and it does not work for long.

 

[Niwot]

Flare gas measurement is probably the most difficult and demanding flow application, especially if there are EPA requirements/standards involved. Ultrasonic and occaisonally, thermal mass meters (depending on the changing molecular weight and gas composition) can handle the extremely low flow rates common in flare lines. However, my understanding relating to VOC compliance is that a measurement device must be in place to handle upset conditions (even if they rarely occur). If you were to take the combination of maximum density, minimum pressure, and maximum POTENTIAL flow rate in a flare header, you'd find that you can reach near sonic velocities during an upset (this can be easily calculated). Ultrasonic meters have a maximum velocity limitation (the signal is carried down the line at extreme velocities). Sometimes, two separate meters are required to handle the extremes of flare gas measurement (from a trickle up to sonic velocities during an upset). Intrusive meters, like averaging pitot tubes can measure the extremely high velocities, but they must be accompanied with a detailed structural analysis showing that they will not break off during upset conditions. They would not work well at normal flare conditions (DP is too low for accurate measurement), but they can cover the high end (upset condition)that would shred most intrusive meters.