positioning of orifice plate 3

[omby]

I have some queries regarding the positioning of an orifice plate in a vertical gas main approx 30 inches in diameter.

Ideally it should be positioned with 10 straight pipe diameters upstream and 5 downstream.
Assuming I only have 10 straight pipe diameters in total (a 90 degree bend at one end and an isolation valve at the other) where would be the best place to position it? I assume that keeping things in the same ratio, it
would be best with 6.6 diameters upstream and 3.3 downstream?

With regards to straight pipe either side of the orifice plate, which would cause the biggest error to the d.p. produced, not enough upstream or downstream straight pipe diameters?

Also how critical is the smoothness of the pipe bore? If the main had at some time in the past been cut through, separated by 6 inches and had a band welded on the outside of the main to join it back together – would this affect the d.p. if it was within a couple of pipe diameters of the orifice plate? Also there is an inspection hatch close by, I assume this would be detrimental too?

I look forward to your replies.

omby

 

[hacksaw]

forget using an orifice plate, there are several insertion d/p and thermal meters to choose from

even if you are using plate flanges, a 30" plate will need special support, etc. especially in thin wall ducting.

 

[omby]

I'm afraid we're stuck with them at present.

Installation of new isolation valves downstream have meant altering the main as they are bigger than those used previously.

The orifice plate installation wasn't ideal to start with, and since I joined the current project (which is not a new installation) half way through and the deadline is rapidly looming, I'm trying to weigh up my options.

I know how it should be installed in an ideal world, but I'm unsure as to the expected size of any errors due to what I have to work with (as in the original post).

 

[hacksaw]

the error and uncertainty are in the 10% range.

 

[omby]

Ok. How do you arrive at that figure?

 

[JimCasey]

Hacksaw is right. An orifice plate will be unreliable at best. And 10 diameters up/5 down, is a very best case installation. Check the ASME tables by Spink. In some cases, such as downstream of a valve, the tables demand over 30 diameters.

Take a look at an averaging pitot tube. Dieterich Standard/ Annubar is one tradename. Meriam Instrument/ Accutube is another. These sample the flow at 4 points across the pipe diameter, and internally average the reading. So you can theoretically read even if the flow profile is distorted. Also, an averaging pitot is easy to install: just a couple of holes opposing each other, with weldolets attached.

 

[JLSeagull]

The key phrase is "fully developed flow profile". ASME, API and ISO publish dimensions. Typically 20 diameters are required upstream depending upon the piping configuration. Flow conditioners permit 10 diamters. Generally, comply with the standards for the published precision / uncertainty. Don't comply with unpredictable precision / uncertainty.

Some head type flow elements such as the McCrometer V-Cone or Fluidic Techniques Flow Pak are available for about 3 diameters upstream.

 

[zdas04]

A few years ago there was a paper written by Mark Haefle and Ron Beatty that was published as an addendum to API 14.3, but I don't remember which section. This paper was a great go-by to develop a numerical value for the affects of any action on total station uncertainty and flow bias and it is worth reading if you can find a copy.

In 30-inch pipe with vertical flow (the OP didn't say if the flow was up or down), unknown pipe roughness, a varying diameter in the upstream pipe, short pipe runs, no flow-conditioner, going through an "isolation valve" (didn't say if that was upstream or downstream or what technology the valve was, it could be a butterfly valve for all we know) or a 90 elbow. The OP also didn't say if he was using an OFU or a cast plate holder (different uncertainties for each). If you run all of these into the uncertainty calculation (I haven't) you would probably get about the same uncertainty as a random number generator and a dead-band so wide that the numbers are simply worthless either for engineering analysis or (god forbid) for custody transfer.

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]

There is a trick to installing an annubar two diameters after an elbow for about 3-5 percent accuracy. However, many unpublished issues would exist. For example, the flange coupling needs to cover most of the pipe wall.

 

[dcasto]

You can get the same low level (+/-10%) accurracy for this installation with a clamp on ultrasonic. Almost any measurement system you put in is going to demand a fully developed flow, no swirls, eddys, off center flow patterns that come with 90's and control valves.

 

[danw2]

Wanna read about flow measurement uncertainties?
Have a go at it, there's over 20 papers available for download . . .

http://www.ceesi.com/pubs_uncertainty.aspx

Dan

 

[danw2]

Or, effects, abnormalities and corrections to oifice plate measurements, specifically

http://www.ceesi.com/pubs_orifice.aspx

 

[omby]

Thanks Dan, those are the sorts of things I'm after.

Jim, I have had experience of averaging pitot tubes and have already considered them but the gas mix in the main can be prone to condensate and I think this may be an issue with such a device.

I know the current installations are not ideal, but they work. I'm aware of how they should be, and there is plenty of info about explaining how they should be, but finding facts to properly explain the consequenses of all the things which 'aren't quite right' is more difficult.

The fact that the installation does work makes it all the more harder to persuade management to adopt a more suitable primary measuring element or even make modifications to the pipework, so for the time being I have to make the best of what I have.

Thanks everyone
omby

 

[zdas04]

When you say the current situation works, how do you define "working"? I've been to a bunch of stations that were as much as 50% off (backwards plate and other problems) and everyone was happy because the chart was tracking. I sometimes think that the definition of "not working" is the pen fell off the Barton.

Have you ever done a material balance using this inadequate equipment? I've gone into some operations where 30% more mass was coming out the back end than went in the front. They wouldn't let me patent the pipe that was making gas. In every case it was lousy measurement.

David

 

[dcasto]

You can calibrate a set of taps on the inlet and outlet of a 90 if you have a true, real, calibrated meter somewhere in the system.
Knowing the estimated errors from the papers will just confirm that the meter is doing what it can do and if you want to use it as a trending device you can't claim a system problem as long as the error is more than the estimated error.

 

[rajant2001]

A flow straightener can be inserted in the upstreme side of the Orifice plate which will improve the accuracy to some extent.

R.Thiyagarajan

 

[fribby]

True a flow straightener would help but if the deadline is looming as Omby stated then it is highly unlikely this is an option given the fab work, breakins required, etc.

 

[omby]

Zdas04, the system 'works' as its part of an air/gas ratio system for a burner which is linked via a dcs to a combustion management system. The system works out gas times, heat required, air/gas ratio's etc and feedback is provided by accumulated heat and rate of temp rise being compared against a known model, plus the analysis of the waste flue gas.
Measurement of air, gas or temperature doesn't have to be far off before poor combustion occurs or a probem is flagged from the supervisory system.

From the initial research I have done I favour using a v-cone meter, though this would be a job for the future, not this project.....

 

[sdb830]

Not sure if you solved your problem, but Emerson sells a conditioning orifice plate that only requires 2 pipe diameters. I have used them for purge air flow and they work great.

http://www.emersonprocess.com/rosemount/products/flow/conditioning_orifice_technology.html