1 reply to this topic

[Once lost now found]

Dear all,

I have been asked to do a kind of sensitivity analysis on the effect that the following three things will have on the accuracy of flow measurement across an orifice:

1. Incorrectly placed pressure tappings.
2. The thickness of the orifice plate.
3. The smoothness (honing) of the upstream pipeline.

The question is rather academic, however, the guys on the plant seem to be seriously considering removing and then redoing all the orifice installations. I want to give them concrete evidence that this isn't a great idea. However they want me to be able to say how much each of the above will effect the measurement accuracies (eg. what would the increased uncertainty be if a flange tap were say 3mm outside of the required 1" distance from the upstream plate face - which is outside the code specs).

The parameters for this problem are as follows: All pressure tappings are flange tappings, the upstream pipes have been smoothed (but not honed), most of the lines are quite small (I don't think there are any lines larger than 6" on the plant). The fluid is helium gas, between 1 and 9 MPa, and 25 and 350 deg C (I know this is a HUGE range, but it covers the entire plant, each orifice won't see this range).

So far I've been trying to find something in literature which would give an indication of the effect that (especially) the incorrect tappings would have on the discharge co-efficient of the orifice. Not much so far though.

If anyone could give me any advice I would very much appreciate that, I'm rather pressed for time as well, so the sooner the better!

Best Regards,

Will

[Andrei]

Will,

The best idea should be to look for the orifice plate specifications from the plant designer, and check if they were installed as prescribed there.
Any orifice measurement has the best accuracy in the second third of the range and decreases exponentially toward the range ends. So make sure that your normal flow falls in the second third of the range. Your measurement errors are amplified by the small density of your gas, so suitable ranges are essential.
From the factors you mentioned, the second one has the most impact on measurement accuracy, it modifies "vena contracta" and the pressure drop as a consequence. But even if you have a wrong thickness, the measurement should work with a proper calibration, the range is affected of course.