Recommended Refinery Flare Header Flow Meters

[ClarkChe]

NSPS Ja is likely causing much more attention to flare gas flow rate than before.

I am looking at the possibility of installing "more than 10" flow meters in various headers in the refinery. I researched back on this forum, talked to many industry contacts, talked with Reps. And still don't have a firm idea regarding which is the "best" flow meter for refinery flare headers.

The criteria are:

Reasonable accuracy, but not necessarily +- 2% (or even better). These meters would be designed mostly to find the flaring culprit(s), not to perform a weight balance on the unit.
Range-ability -- flow rate, composition, temperature, and pressure will vary.
Reliability – obvious
Cost – we are considering installing possibly 15 of these in our refinery, so cost is a major factor.

Based on literature, the FCI (Fluid Components International) ST100 looks good, and cost is reasonable. This meter is a mass dispersion meter, and is affected by composition change in the gas. But it also has 5 calibration points, which conceivably could be part of a flare rate correction for varying compositions.

If others have fairly recent experience with flare gas meters, or recommendations, or some useful insight, please do reply back.

Thanks you.

Clark

 

[pennpiper]

You should consider the FloTap meter shown here:

http://www.metrontechnology.com/pitot.htm

They only require a small connection which can be installed via a Hot-Tap. This would be done without a shut-down. The meter can also be inserted and withdrawn as required without shut down. This would allow you to reduce cost by installing only the 15 connection points but purchase only one meter and move it from place instead of purchasing 15 meters.

prognosis: Lead or Lag

 

[poli60]

We used ultrasonic flowmeters (see the attached paper).
No pressure drops, no troubles in case of high temperatures... it was a kind of "forced standard" for us.
I'm not updated about the developments of last five years.
good luck!

 

[sshep]

Hey Clark,

At one of my previous sites we used an ultrasonic flow meter by panametrics which looks to be as per poli60's artical. It was installed in about 2009. Because it involved a platform to access the instruments, it cost more than was originally intended, but the transmitter (where all the action is) can actually be located at grade with cabling to the transducers.

The main problem with our purpose was we wanted something more accurate than was possible with varying MW. There is some correlation of soundspeed with hydrocarbon MW which the rep supplied in a paper on the subject, but N2 (which we used for guard bed regen) and hydrogen (from hydrotreater purges) are exceptions and threw this off. By running extra wires from the transmitter (which we should have done but didn't), it would have been possible to get soundspeed and other useful diagnostic data. Don't forget to pull those extra wires if other info is available from the meter. The panametrics rep had to come out a few times to refine the parameters in the meter which was done by either computer (probably) or hand held device, I can't remember anymore.

By making suitable assumptions it eventually worked good enough to declare a victory, but we had really hoped for something better. Fortunately it sounds like your needs are more of a "flow" or "no flow" application, so hopefully it will be easier.

best wishes always,
sshep

 

[ClarkChe]

pennpiper
poli60
sshep

I appreciate all of your comments and information.

We currently have a Panametrics on the total flare. I am not critical of the Panametrics, but we are looking at possible other meters for the 15 or so flare headers in the refinery.

This is not really a "flow vs no-flow" application. In a best world, I would like to be able to add up the individual flare header rates and come reasonably close to the total rate. Yes, that does seem more of a weight balance, but not truly. I just would like to have enough confidence in all my flare meters, so that I know where the individual flare gases come from.

As I mentioned before, we are evaluating the FCI ST100. It is very good on paper. If interested, please look here.

[URL unfurl="true"]http://www.fluidcomponents.com/Industrial/Products/MassFlowMeters/ST100-Series-Gas-Mass-Flow-Meters.asp[/url]

The rep does say that constant composition gives very accurate results. A flare does not guarantee constant composition, so we are working with the rep to determine FCI's best estimate of how accuracy is affected by what we believe will be the composition changes when flare rate changes. FCI has an evaluation model that does that.

I was also hoping that others had gone this course, and already had results on this study.

Thanks,

Clark