Natural Gas Flow 1

[Cat3347]

I have a Natural Gas flow application. Upper range flow is approx. 95000 SCFH in a 3" line. Have used turbine meters, but they do not hold up well. The only area to mount is an 8' vertical line (flow downward). Would a flow nozzle be a good fit? Line pressure is around 15psig.

 

[JLSeagull]

Consider a V-cone from McCrometer.

 

[Ashereng]

I just installed an annubar in a gas pipeline service at 10 psig supply pressure. Maybe that would work for you too.

We also looked at a thermal mass meter. That also worked for us, but I guess we don't use a lot of those so spare parts may be a bit of a worry.

I haven't used v-cones at all, but it should also work.


On a slightly separate note, here is an interesting site:

http://www.flowmeters.com/ufm/index.cfm?task=main

Hope it helps.

"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?

 

[nbucska]

Before we can start to design we should read the specification.


Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[Ashereng]

Who's designing?

These are thus far simply suggestions of possibilities with respect to technology.

"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?

 

[dcasto]

I can list 10 technologies to meter the gas. What everyone needs is an explaination of desired accurracy and what may have caused the turbine meter to fail.

An orifice meter and its clones anubar, pitot tubes, venturi, ect are simple and easy. but suffer from accuracy at high Dp/P ratios. If you can live with a 3% to 5% error, they are very easlily maintained and inexpensive. Higher accuracy, ultrasonic meters, but very expensive and exotic to maintain.

mass meters (microMotion etal) very accurate high pressure loss, no way to calibrate in house.

 

[Ashereng]

dcasto,

The OP stated 15 psig line pressure, which is relatively low pressure. Hence, my example are from low pressure applications (10 psig) that I recently investigated and selected.

I am giving suggestions related to low pressure in the line.



"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?

 

[JLSeagull]

Flow meter selection often amounts to limiting the technologies for the application. I am not a big "fan" of turbine meters. (The pun was not planned.) If they don't hold up, select something else.

The 8 NPS line is really big for a Coriolis meter. Micro Motion lacks an 8-inch meter.

Head type meters require some pressure drop and you have little pressure. The averaging pitot tubes like Annubar and the V-cone should drop less than an orifice plate. Run the calculation to see if a head type meter works with the pressure. Turndown is another issue but less significant these days with the ultra-accurate dp transmitters.

Transit-time ultrasonic could be OK, perhaps better than 2 % accuracy. Thermal mass flow is not accurate (worse than 5 %) but often work in high turndown like flare headers.

Accuracy is less of an issue for internal unit flow measurement. However accuracy or precision is significant if for custody transfer. This can even apply for meters for accounting between process units.

 

[Cat3347]

Thaks for the replies...

Some notes for consideration:

1. Accuracy needs to be better than +/- 2%
2. We need a compensated flow (temp and pressure)
3. The turbine meter is failing due to design. Velocity?
4. I have sized the application for an orifice plate (

http://www.flowexpertpro.com)

and it checks out fine, but has anyone had experience with verticle-downward gas flows and orifice plates?
5. I also looked at an averaging pitot tube, but could not get the dP high enough to feel comfortable with it.

 

[dcasto]

unless you have a realtime gas chromatagraph, forget the 2% with anything other than a mass meter.

 

[joecontrol]

I think there are positive displacement units out there that could that flow and pressure with very good accuracy. I would suggest getting a pulser output send it to a recorder and do the math in the recorder.
This is not cheap but i have done it before and have had good results.

Joe