7 replies to this topic

[dhns]

Dear sirs,

Am looking for some solution to measure my oil flow in the multiphase pipelines (except MFM and Test Separator).
The pipeline sizes are 12’’ to 36’’, we are thinking about the Coriolis mass flow meter to measrue the flow. I am expecting the Coriolis meter is the best one to measure the mass flow of the liquids. Can we use the Coriolis meter for multiphase flow measurement? (our plant production fluid contains ( <1 volume% gas) water cut is varying 1 to 100 % with respect to oil flow.
Please advise the experienced people on the above regards.

Thanking you
dhns

[ankur2061]

dhns,

I found this by googling and apparently this answers your question:

http://www.documenta...s/mc-001022.pdf

http://www.documenta...s/ar-001154.pdf


Regards,
Ankur.

[JMW]

A lot of work has been done by some of the major coriolis meter companies to address the problem of multi-phase flow and the obvious attraction of upto a million potential well head metering applications.....
Both Micromotion and Invensys have done a lot of work on two phase flow that I know of, and both companies should be consulted.

But, the problem is the range of different two phase (liquid and gas) flow regimes that can exist (and the problems of slip i.e. where one fluid flows faster than the other) and it seems coriolis can be successful with dispersed bubble flows but beyond that they are less effective.

I would recommend visiting the Neftemer site and reading some of the papers given at conferences. e.g. http://www.neftemer....on-tog-2008.pdf which contains a diagram showing the range of application of the technologies.

But if your entrained gas is less than 1%, this is within their capabilities.... the question is, how much less than 1% because at some point you may not need the entrained gas capability and this you may need to discuss.

Edited by JMW, 13 June 2011 - 05:14 AM.

[kkala]

I do not have experience on two phase flow measurement, but Chemical Engineering, Jul 2008, Newfront "Go with the flow" writes for coriolis meters.
-The two largest advancements in coriolis technology from Micro Motion include a "dramatically improved ability to measure mass flow accurately in two-phase flow conditions and the ability to check or verify the meter's calibration in-situ", says O' Banion.
-The two-phase flow ability improves measurement during periods of continuous bubbles in liquid, as well as the ubiquitous situation in the CPI batching from empty. This is where users start with an empty pipe, fill the pipe while taking measurements and then end with a purge cycle.
Note: O'Banion is director of Micro Motion in global chemical industry marketing.
Consequently above seems to agree to previous post by JMW; nevertheless tendency for coriolis meters is to get wider and wider range of applications.

Edited by kkala, 19 June 2011 - 02:31 AM.

[JMW]

There are some different stories coming from the coriolis people. Well, these may be different aspects of the same feature.

In one it is said that the gas capability is a benefit of low operating frequencies where the velocity of sound effect is minimal.
Certainly the VOS effect is recognised as affecting vibrating element sensor density measurement and the EGA (Entrained Gas Amplifier) Solartron density meter, now part of the MicroMotion product range, is capable of 0-100% entrained gas by virtue of driving at an alternate harmonic (two nodes) of the resonant frequency. The price is lower accuracy but results stable repeatable and accurate readings even though that accuracy is less than for 0% GVF and operating at the usual harmonic (three nodes). There is more to it than just VOS, there is the ability of gas to flow through the liquid which means that in low viscosity fluids the gas can tend to migrate to the tube walls so that the motion of the tube only displaces gas and the liquid flows axially through the fluid.... and hence the tube. The physical movement of the tube is very slight.

But Mass is a function of the phase angle, not the resonant frequency, the way the tube behaves differently along its length, so though related, the problem is more complex.
The EGA density meter can handle bubble flow, slug flow and even, when in vertical alignment, it can be used as an interface/level device with one end of thee tube in liquid and the other in gas. But these are significant effects to handle for mass flow.
Imagine that a gas pocket moves through a tube. The mass rate of flow varies along the length of the tube because the mass of the gas is so much less than the mass of the liquid. But bubbles dispersed in the fluid result in a reasonably homogenous fluid. Note that this is specifically commented on in the MicroMotion blog - there is a discussion of the effect and how it is handled here: http://community.mic...n.com/group/20/
But don't be too surprised that they don't explain everything.
Invensys may or may not address the problem the same way.... I don't know.

Note that the current limitations for mass flow need not be too restricting.
One of the problems, especially with more viscous fluids, is the difficulty separating gas and liquid. It is reasonable to separate out large amounts of gas but to leave a proportion of gas - as bubbles, still in the fluid. This was a problem for coriolis and other measurements. So the resolution of the problem is either to improve separation technology, which the Auger referred to does, and/or improve the meter's ability to handle the residual gas in the liquid...
For well metering, the coriolis meter only currently is capable of use in GVF 0-10% but the problem is well production quality changes with time.
Well metering is a problem for lots of technologies, not just coriolis. The Neftemer presentations show that MPM of all technologies account for a very small percentage of the total opportunities. If any technology can cater to a wide range of GVF and deliver good accuracy then there is a chance to clean up. The size of the market is huge. This warrants a lot of investment, if there is a chance of success, and it means the coriolis meter people will look at any opportunities that present as a means to recover some of that investment and learn more in any other industries or applications.
So as Kkala says, the expectation is that coriolis will get better because there is a suitable incentive to justify the investment.
The question is, where is the boundary? There may be a technical limit that means no matter how much money you invest, you are only incrementally approaching a limiting boundary. It might be that coriolis can handle dispersed bubble flows i.e. reasonably homogeneous fluids, but that's it. Maybe they can find a way round that. Who knows?
Essentially the coriolis meter is a mechanical device. The measurement is the interaction of the vibrating tube with the fluid it contains. Possibly radiation or ultrasonic techniques have a better potential. For example, the improvements in Uultrasonic meters have been in time of flight cross pipe measurements and multichord technology to derive the flow profile or to respond to changing flow profiles, if you like. If the problem with coriolis is the different mass flow rates along the length of the vibrating tube, then maybe ultrasound techniques can adapt by sensing how flow varies along the length. (and ultrasound is measuring volumetric flow, not mass and this varies far less - it is the density that really changes).
I don't know the answers here, I have been away from flow measurement for some time now.Ultrasonic methods are certainly available for GVF measurement.
Work on profiling cross pipe flow for gas and liquid distribution is a feature of downhole logging in horizontal wells so we can expect this to be another area rich in investment that may deliver some benefits to other industries.

Edited by JMW, 19 June 2011 - 08:02 AM.

[kkala]

Picture of actual situation of coriolis flow meters is given thanks to the post by JMW. I have detected their use here in the fuel oil to refinery steam boilers or to cement rotary kilns.
Maintenance of coriolis meters, if needed, have to be realized in specialized shops out of the Refinery.
Refinery has recently considered them expensive, so positive displacement flow meters have been specified instead of them for a new boiler.

Edited by kkala, 19 June 2011 - 02:59 PM.

[JMW]

A couple more comments.
There is a distinction between air pockets in the flow and air bubbles dispersed in the flow. In the latter case the resultant fluid may be relatively homogeneous.
Air pockets cause problems for most flowmeter types - it results in false registration and it may also cause damage.
With thin fluids it is usual to use air separators which vent the air from the flow stream just prior to the meter. In the case of LPG meters there may also be a pressure regulated back pressure or shut off valve downstream of the meter.
With meters equipped with electronic registers, it is feasible to inhibit meter registration when air pockets are detected. This could be by using vibrating element density or viscosity sensors or the density signal from the mass meter. Optionally it can be a GVF measurement device e.g. ultrasonic. http://www.cidra.com...cts/gvf100.html
This is fine if the air/gas pockets result from starting the lines empty, purging the lines at the end of operations or when switching tanks when one tank empties. In bunkering as each tank empties the pumps attempt to scavange the last remaining fuel in the tanks. This draws in air too. In trials of the entrained gas coriolis meters in Singapore the operators deferred tank stripping, as it is called, till the end of batching because the assoicated errors become significant if the incidence of entrained gas increases. In my experience of coriolis meters (some years back now) the onset of entrained air can give rise to very significant measurement errors. If the meter's own density signal is used to detect the entrained air and inhibit further registration, it can only do so once some corrupted data has been registered. If this is the same issue for modern coriolis meters then a better approach might be to use air detection sensors upstream of the meter to detect the air as it reaches the sensor such that they can inhibit registration before the air/gas can reach the meter. A time delay when the air has passed the sensor would mean the mass meter wouldn't start to register again until the air had also passed the mass meter. This means that not only the fluid carried over with gas pockets is not registered but also some fluid without gas pockets on the leading and trailing edges of the air pocket.
Another approach is to use volumetric meters. The advantage of a PD meter, for example, is that its measurement is not corrupted. It accurately reports the volume of fluid no matter that it has a significant amount of gas in it. Of course, the problem then is that the proportion of the volume that is fluid is not known. But, if a GVF sensor is used, it can be used to compensate the electronic registration of the volume. So if it indicates 50% gas then the registered volume is reduced by 50%. Another option is to use the EGA density meter which, together with the volume measurement will report mass flow. This is independent of the density of the gas or the liquid. It assumes that the mass of the gas is insignificant (this is the same assumption that coriolis meters make when dealing with bubble flow... what they actually measure is the mass flow of the liquid plus gas).
This approach requires choosing a suitable PD meter. Not all are mechanically suitable to handle air pockets and can suffer damage or excessive wear. Others, however, are relatively unaffected.

[JMW]

There is a link in the Cidras website to entrained gas flows: http://www.cidra.com...g_products.html