Ejector effect in two joining pipelines? 1

[DAT17]

I have a problem with flow measurument in two joining water lines that seems to be caused by an ejector effect at the connection. The main header is a 10 inch line with normal velocity of 1.75 m/s. I then have a 6 inch line connecting via a reducing T. After the connection the line velocity in the 10 inch header is 2.44 m/s. We hve tested the flowmeter and am sure that the instrument is fine. The line is not under vacuum but we have a suspicion that the 6 inch line might be - due to an "ejector effect" caused by the larger line and flow.

Does anyone have a similar experience with a piping arrangement like this in a seawater system?

 

[BigInch]

You don't explain what problem you're having.

the 6 inch line might be - due to an "ejector effect"

doesn't mean anything I can understand.

All I see is that you have a flowrate in the 10" of 319 m3/h before the tee and you have a flowrate in the 10" of 445 m3/h after the tee, so I would assume the 6" connection is adding 125 m3/h to the 10" pipeline.

Are you saying the 6" is not adding 125 m3/h to the 10" pipeline?



**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[DAT17]

Ah- thanks for that BigInch

- the problem is that I know I am getting the flow of 125 m3/h, but the flow meter is jumping around the measurement point - I have fluctuations between 70 to 150 m3/h. We have checked the grounding on the instrument and this is fine, i also do not believe the instrument is at fault, because at low flow (below 60 m3/h) the flow is steady and not variable at all. When the flow is increased as above I have this problem. I also have the correct pipe diameters upstream and downstream of the flow meter (a magflow).

Hence my initial thinking that the main flow might be creating a slight "ejector effect" (or sucking) which then impacts the flow measurement.

I'm also looking at the piping layout to see if this could impact the smaller line's flow (e.g. make it more erratic). The 10" flow connects into a T-piece (instead of a 90" bend) with the reducing T right next to it where the 6" line ties in. I have a suspicion that the line turbulence caused by the flow passing into the 1st 10" T may adversely affect the flow from the 6" line, which in turn transfers back to the flow meter in question. This sound feasible?

 

[BigInch]

Turbulence can cause many unusual effects and if you are sure about your meter calibration, turbulence would be the next suspect (usually its the first). I can't quite picture your arrangement, other than confusing, so that's another signal for turbulence problems. Can you post a diagram (see upload your file to ENGINEERING.COM below).with dimensions. The typical upstream and downstream lengths for meters assume there are no overly complicated flow patterns set up nearby. Those are minimum recommended distances from mostly open valves and straightening vanes etc. in pipes of uniform diameter, not really for complicated reducing-T connections.

With reducers and reducing tee connections, there are additional pressure waves that reflect back from each change in cross-sectional area, which are actually responsible for accelerating or decelerating the flow as it moves across the change in cross section, so there is a greatly increased additional potential to develop turbulence when reducing fittings are nearby.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[zdas04]

Regardless of what your measurement is showing you, the main flow path through a tee cannot "suck" on the branch like you would see in an Eductor. The eductor works at the no-flow boundary between the power fluid and the suction fluid and this boundary is very limited in a 10X6 reducing tee.

I've seen flow-visualization demonstrations of the flow into a tee and the branch flow retains part of its identity for a considerable distance (certainly dozens of pipe diameters) downstream of the tee. The distance is a function of the magnitude of the difference in SG, temperature, and specific contaminants. As long as the two streams retain some of their individual characteristics, random perturbations in the flow will be significant. As soon as the two streams are homogenized into one stream the magnitude and frequency of the perturbations will decrease.

A tee is not an eductor or ejector. You can get a "pop bottle effect" (i.e., the sound that a bottle can make when you blow across its neck) in a tee that is partially full and at high velocity (think of the pipe running off a cliff and the tee towards the bottom), but the amount of pressure reduction is miniscule.

David

 

[DAT17]

thanks David,
sounds like I can discount the ejector effect, but what you said about the "pop bottle" effect sounds worth investigating.

BigInch,

see enclosed the arrangement I have at present, the flow meter is 2D from the elbow.

We are thinking of taking the 6" line for a walk and move the flow meter to the vertical, this will increase the distance between the reducing T and the flow meter to around 3 to 5 metres, which should help if we have any pulsations in the line. This would also inlcude a 180 deg turn which would probably help to dampen any pulses.

 

[DAT17]

Typical....
forgot to complete the upload, here's the rough sketch.

 

[BigInch]

I can't find the diagram. No worries. 2D is what I didn't want to hear.

I would definitely increase the distance from the meter to the connection. There is no way IMO that 2D is sufficient for anything but laminar flow, even then I would still have doubts.

If you see pressure fluctuations as severe as you say they are at the meter, that should be proof enough that some significant turbulence is going on there.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[DAT17]

Thanks BigInch and Dave,

Think we will drop the theory of the ejector and focus on getting the flow meter away from the bend!

 

[BigInch]

My eyeball CFD can see major problems with that. Within 1/2 meter you have two significant opposing flows trying to bounce off the opposite walls while one tries to get past the other one while they both speed up as they finally exit right. Turbulence cubed. Those swirls i would imagine break off and flow past, every time leaving a large pressure pulse. I do think you are having a hard time maintaining the same pressure at the outlet of the bend with all those swirls passing by it, hence not a constant outlet pressure at the bend, nor a constant flowrate through the meter.



**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[BigInch]

And that's only 2D.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[DAT17]

Spoke to the guys on site and we agree with your comments, we're definitely going to change the inlet 6" pipe, we can't do much about the 10" T's, but we have a lot of flexibility with running the 6" line, so we're going for max D between the flow meter and the bend!

 

[BigInch]

Great. Maybe a check valve in that 6" would help too. Can you fit one in before the T?

Hope it works out.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/