Orifice Plate Turndown 1

[younginstengr]

Everything I find online or otherwise states that orifice plates have a turndown of 3:1 for 1% uncertainty, 4:1 for 2% uncertainty, and 10:1 for 5-10% uncertainty. The DP transmitters we typically use have a turndown of 40:1.

It seems to me that the turndown of the transmitter itself would be the limiting factor of the flow measurement. Why is the accuracy of orifice flow meters confined to such low turndowns when there are DP instruments capable of much larger turndowns? Is this due to a limitation of the curve produced from the orifice plate itself?

I'm new to the field, and any help would be appreciated.

 

[JLSeagull]

Orifice plates are primative devices. The piping upstream and downstream is not round. The plate is not centered. The sharp edge has burrs. Upstream devices cause swirl. Regardless of the transmitter accuracy the flow rate derived from pressure differential lacks the sort of precision that you could expect from something like a coreolis meter.

API MPMS 14.3.1 Table 1-7 shows an uncertainty estimate for a natural gas flow calculation. Sort through the MPMS flow data and develop your own opinion then share your thoughts.

However, if an application had suitable turndown with a half percent transmitter, the accuracy of the same installation would be better with a 0.05% transmitter.

 

[hacksaw]

your uncertainty data is based on practical meter accuracy in operating plants. JLS is right to point out that a lot of factors are involved.

 

[SaskGear]

The main reason for the low turndown is the relation between flow and dP not being linear. Volumetric flow is based on the square root of the dP, which means that your 40:1 transmitter is really only like 7:1. Even with that the low end of the transmitter is not as accurate, most transmitter errors are based on % of full scale so as you get close to the low end of the range that % error at each reading gets larger, so you could take a couple more points off of the turndown for that reason as well.

 

[JLSeagull]

YES, what SaskGear said.

In contrast, some other flow technologies quote accuracy in percentage of flow, not percent full scale. This is quite different too.

 

[JimCasey]

The signal generated by the orifice plate is proportional to the square of the flowrate. So the very, very, best you could hope for with a 40:1 transmitter is a little over 6:1 ( 6.32 = Sqrt 40)

Meanwhile other flowmeters such as Magmeters or Coriolis meters have 100:1 turndown and are more stable and robust than orifice plates.

There ARE places where the orifice plate is the best flow element for an application......I just can't think of any.

 

[hacksaw]

as long as we are talking about low viscosity fluids and the meter dp is a small fraction of the upstream pressure in compressible flows i agree, but many of my meter satisfied none of those criteria on extended range and we were lucky to get 2-3%

the really high precision flows require a lot more than just good xmtr turndown

 

[JLSeagull]

The orifice plate is a good flow element when your oil contract for custody transfer requires a seven-day windup chart at a location with no power. Most of those places are upgrading to wireless and solar power.

 

[younginstengr]

Thank you for your help, especially you, SaskGear.

 

[dcasto]

let me add that you can have stacked DP's and get a much wider range with high accuracy.

In the US ethylene is measured with such systems and at values 10 times that of oil or natural gas, the measurement reliability and accuracy is very important.

 

[Ulon]

Other than Stacked DP, Dual Span DP is also available now. The DCS will contol the span selection.