Orifice Meter Max Differential Pressure 1

[GBinns]

What is the maximum recommended differential pressure across an orifice meter run in a lean natural gas service?

Normally the meter would be sized for around 75" H2O so you can use a 0-100" H2O transmitter. What if you have an existing meter run and don't want to change anything, but you want to flow a lot more gas? Can you just re-range your transmitter to read upwards of 2000" H2O and howl the gas through it (assuming you have the inlet pressure required)? Is a bent/eroded plate the worst thing that could happen? What would the symptoms be of a bent/eroded orifice plate? What happens to the meter accuracy?

 

[zdas04]

Could you give us more information? Like Line pressure, meter size, max beta ratio you'll accept, and are there any flow conditioners.

If you have a 2-inch meter run with a 1.5 plate, you can "measure" upwards of 24 MMCF/d at 1,000 psig and 2000 inH2O at 80F, but your velocity is 190 ft/sec and you are going to scour any passivation layer off the steel.

Thinking back on the assumptions that went into the development of the equations, I'd think that as long as the plate didn't bow you'd be good up to about 0.3 Mach, (somewhere around 300 ft/sec depending on specific fluids and pressure/temperature). If the plate bows then you'll be reporting low by something on the order of 15% (to say nothing of the difficulty of removing the plate from the fitting). I can't remember a specific Reynolds Number upper limit (the example above is over 15 million), but there may be one.



David Simpson, PE
MuleShoe Engineering

http://www.muleshoe-eng.com

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

The harder I work, the luckier I seem

 

[JLSeagull]

maximum recommended differential
Study the plate thickness selections in the API standards and find that the differential pressure is capped at 1000 inches water.

For gas applications a rule of thumb limits the pressure differential in inches water to the inlet pressure in pounds per square inch. You don't want the meter to drop a significant percentage of the available pressure just to measure the flow rate.

The normally sizing for 100-inch water dp accommodated the 100 inch water mercury manometers. You can buy a smart dp capable of 1000 inch water and select 0-50 inches with suitable accuracy.

If you have an existing meter run you can increase the orifice bore toward a beta ratio of bore and pipe inside diameter toward 0.7.

Other technologies exist. An external clamp-on transit-time (time of flight) ultrasonic meter may be suitable for some non-custody applications.

Apporacing 2000 inches water will bend a large plate. Check into the API MPMS for guidelines. Gas flow is covered in 14.3. API Chapter 14 “Natural Gas Fluids Measurement,” Section 3, “Concentric, Square-Edged
Orifice Meters,” Part 3, “Natural Gas Applications” of the Manual of Petroleum Measurement Standards, Third Edition,
August 1992.

 

[JimCasey]

If you are using an orifice as a flowmeter, a lot of the sizing equations break down outside of the normally encountered ranges. Size the plate for a 0.7 beta at your anticipated flowrate for metering.

If you have an upset condition that causes an offscale high flowrate, I doubt if you would create enough DP to hurt the plate. If you are worried about that, 1/4" thick plates are available. Where you might be concerned is if the high flow picks up particles in the system and the edges of the orifice are eroded. For accuracy, the edges of the orifice must be very sharp and uniform, not eroded or dinged by chunks of weld slag.

DP instruments are almost universally protected against overrange. You are not likely to hurt the transmitter with a flow surge.

 

[Ashereng]

We standardize on 1/4" orifice plates as minimum thickness.

Orifice plates are not expensive. Could you not change the bore to suit your new flow conditions?

Or, are you needing to change online without breaking the line? In that case, would a senior or rotary changer work?

"Do not worry about your problems with mathematics, I assure you mine are far greater."
Albert Einstein
Have you read FAQ731-376 to make the best use of Eng-Tips Forums?

 

[slacey54]

If this a not a custody transfer meter you can go to 200" or above. BUT - you loose accuracy. That is why most custody transfer contracts are written to current AGA regulations which are normally a max of 100" w.c. Knowledable buyers and sellers both actually prefer flows 20 to 80 w.c. Accuracy suffers above and below these limits when trying to maintain an accuracy of plus or minus of less that 2%. Accuracy can suffer considerable either way when even one small drop of fluid or small (pin size) gets on the upsteam side of the plate or fluid or debris on the upstream or downstream ID of the meter tube. The accuracy suffers even if there is something attached on the downstream side of the plate. As far as overranging you delta P TX just refer to the max flow spec's. They are quite versatile. As far a damage is depends if you have a junior or senior meter run and what equipment if any is downstream of the meter run. I have personally seen over 100 bowed plates. It is actually very common at well site delivers. It does happen at custody transfer meters but not as often. Bowed plates can be caused by slugs, control valve malfunction, SCADA sytem operator error, non calibrated and tuned SCADA or pneumatic contolled controllers, etc.. A bowed plate, or worse yet, a dislodged plate can be very time consuming and expensive to get extract due to damage to the meter run or equipment downstream. Obviously the plate itself could cause potential catastrophic damage to anything downstream of the meter run itself, i.e. straighting vanes, compressors, turbine meter, control valve, separator dump valves, etc...

You just have to ask yourself what is you gas accuracy is worth to you and any related down down for plate retraction.

 

[cpierson]

As you go to higher pressure drops, the net expansion factor, Y, in the meter factor will need to be changed to keep some accuracy (per Crane TP 410 Appendix A21).

At that extemely high of pressure drop (2000 iWC = 72.22 psig) that you are talking about, I would think that you would reach sonic velocity (but depending on your inlet pressure). At that point all accuracy is shot, because the orifice plate will act as a choke (i.e. decreasing downstream pressure will not increase flow).

 

[dcasto]

There is also an overlooked ratio that must be watched, the ratio of dp/P all the data that was used to build the equations for flow were within in a range of this ratio. We really saw this with vacuum gas gathering we wanted to have 20" as proposed, but 20"/5psia was outside the limits of the equation. We increased the orifice and dropped to 10" and our measurement improved.

The 20" to 80" limit is only for paper chart integration and it is really an error caused by the act of integrating the charts. If you read the transmitter specifications they state accuracy (keep it simple here no repeatability, bias, ect analysis) as % of range so it typically .15% of range, so you should always try to run as high in the range as possible for the least amount of error.

 

[maddocks]

Most gas receivers cap the maximum differential at 200" wc.