Reverse installed orifice plates 2

[AntonTRN]

Unfortunetaly several square edge orifice plates are installed reverse. (upstream / downstream), so the sharp edge is upstream. What is the effect we can expect, less delta P? What kind of extra unaccuracy we can expect?
I have checked the Shell flowmeter handbook, but I can not find anything about this effect.

 

[zdas04]

The literature uses numbers between 24% and 28% low. I've done some experiments that support that range and have had a couple of meter stations over the years that were in that range of error.

Of course, anyone who puts a plate in backwards might be prone to other errors (like not checking for a nicked edge, not getting the seal on righ, damaging the seal, etc.) and the station may actually have a higher cumulative error.

David

 

[JLSeagull]

Normally orifice plates are marked with the word upstream. However, the piping normally is not marked with the the word upstream or the flow direction. Someone with knowledge of the system or a P&ID might not make the mistake. However, the installation drawings with the construction craft person often lack sufficient detail.

 

[zdas04]

There's always an indication. Sample ports are installed downstream, check valves have arrows, etc. People make mistakes no matter how hard you try to prevent it.

I had a guy install both plates backward on a 12 MMCF/d well with two 6-inch meter tubes. 3 months later when the problem was found we got a pretty big sales adjustment. I went to a meeting with my boss and the head of gas measurement a few days later and this came up. The measurement ditz said "how could any idiot put a plate in backwards?" so I drew a cross-section of a plate on the board and asked her to tell me what the flow direction was. She said "left to right, the bevel makes it easier for the gas to enter the hole". My boss said "they call it 'square edged orifice measurement' for a reason". Even "measurement professionals" with Engineering degrees can get it wrong.

David

 

[danw2]

 

[danw2]

This thread on control.com has some web references for errors associated with orifice plate mis-installation:

http://www.control.com/thread/1241201727

Dan

 

[AntonTRN]

Thanks a lot!
I try the formula %DQ = -18.93 + 12.91b - 34.04 (E/Dnom) – 8.900 (e/E) + 13.64 (e/E)2 but it leads to crazy figures. (Morrow and George have proposed the following correlation for a backwards-facing orifice plate to estimate the percentage measurement error as a function of the beta ratio, b, plate thickness to meter tube diameter ratio, E/Dnom, and bore thickness to plate thickness ratio, e/E.)

To my unnderstanding it is the factor e/E which blows up.
Has anybody an positive experienced formula?

 

[hacksaw]

gee, the disch coef for a conical inlet plate is about 0.734, you can try that

 

[roydm]

Ahh, why not simply turn them around?

 

[ccfowler]

Am I missing something? The initial posting phrase, "the sharp edge is upstream" suggests that the orifice plates were installed properly.

I've encountered several cases where orifice plates were incorrectly stamped as to which face should be upstream, and as a result, they were improperly installed. The installation errors were always found as the result of calibration data that didn't make proper sense.

 

[zdas04]

I think we all missed the same phrase. I think we've all been telling him how to fix something that isn't broken.

David

 

[AntonTRN]

Indeed, the fact is we all know how to install to stamp and mark orifices, At this moment we have several pieces installed wrongly. They have to sit there the next 2 years, what I have to find out is how much deviation we have an if it is lineair over the range. At this moment I expect -20%.
Does anybody has the facilities to do a test, if required I will sponsor by sending an square edge orifice ?

 

[danw2]

Both Morrow and George appear to still be working for or at SWRI, Morrow as recently as February; George in April (dates of published papers).

Have you contacted either guy or SWRI itself ?

http://www.swri.org/contactinfo.htm

(210) 684-5111

From the paper on orifice error:

Dr. Thomas B. Morrow is a technical consultant at the
Southwest Research Institute, with nearly 25 years of
experience at SwRI. During his tenure at SwRI, he has been
involved in many research programs for the oil and gas
industry and the U.S. Coast Guard, among others. He is
currently the manager of ultrasonic flowmeter research and
energy rate meter research at the Gas Technology Institute
Metering Research Facility at SwRI. He is also the past
manager of orifice flowmeter research. Dr. Morrow has
contributed significantly to the natural gas measurement
standards developed by the American Petroleum Institute,
the American Gas Association and the International
Standards Organization. Dr. Morrow received his doctorate
in mechanical engineering from Stanford University. He was
also a British Railways Research Fellow in the Department of
Transport Technology at the Loughborough University of
Technology in England. He has prior work experience at E.
I. DuPont de Nemours & Co. Dr. Morrow has written over 50
technical papers and is a registered professional engineer in Virginia and in Texas.

Dr. Darin L. George has been a senior research engineer at
the Southwest Research Institute since joining the staff in
2000. He is currently the manager of turbine flowmeter
research and natural gas sampling methods research at the
Gas Technology Institute Metering Research Facility at
SwRI. Prior to joining SwRI, Dr. George worked as an
engineer and reactor reload analyst at the U.S.
government's Savannah River Site and, also, as an engineer
at Sandia National Laboratories. He received his doctorate
in mechanical engineering from the University of Michigan.
Dr. George is a member of the Multiphase Flow Committee

 

[AntonTRN]

Thans Danw2, I have contacted the authors and been helped very kindly. I have put the figures in a spreadsheet; the deviation is between 11 and 15% for orifices between 2 and 10".

Thanks all for your help.