This article was printed in CONTROL's July 2009 edition.
“Ask the Experts” is moderated by Béla Lipták, process control consultant and editor of the Instrument Engineer’s Handbook (IEH). Preparation of the next edition of Volume 3, “Process Software and Networks,” is in progress. If you’re qualified to update an existing or to prepare a new chapter or would like to participate in this column, write to email@example.com.
Q: I have to select a flow transmitter that can measure flow within a range of 5-150 m3/hr with ±0.5 full-scale accuracy. This flowmeter is used to generate a feedback signal for a control valve that controls the flow of water to an ejector, serving to accurately maintain the density by dosing a chemical into this ejector. Is a Coriolis flowmeter is a good selection for this service?
I have studied about many flowmeters, such as orifice plate, vortex, etc., and most of them have some non-linearity, low rangeability and many other problems. In the case of the Coriolis meter, I think I can use it, but I am not sure.
Other features required are linearity and low maintenance because this section of the plant is critically important and must be continuously available.
Any advice you can give me would be appreciated.
A: Coriolis is a good mass flowmeter. It can provide the 30:1 rangeability you need on a mass flow basis. If you need volumetric flow detection, then temperature compensation is needed. You should double-check to see if you can live with a 0.5% full-scale error, because if the full-scale reading of the meter is 150 m3/hr, this error corresponds to an absolute error of 0.75 m3/hr. At your minimum flow of 5 m3/hr, this error corresponds to 15% of the actual flow. and it might not give you the accuracy needed in your density to flow cascade loop. Some manufacturers claim rangeabilities as high as 100:1 and errors as low as 0.1% FS, but these claims—in my experience—are exaggerated.
You could also consider vortex, turbine and magnetic flowmeters, insofar as they can provide volumetric and linear flow measurements and can also be accurate to 0.5% FS error or better. These meters cost less than a temperature- compensated Coriolis, but the 30:1 rangeability is too much for them. If you can relax that rangeability requirement, the magnetic flowmeter is a good choice on water service. (Some magnetic flowmeter suppliers claim their units meet the 30:1 requirement, but my experience contradicts that.)
Therefore, if you want to accurately control the density at the ejector outlet, you should use a density controller as the cascade master of the system. The output of this controller should provide the set point for the slave controller, which modulates the flow of either the dosing chemical or water. By the way, when we are mixing two liquids to obtain the desired mixture density, we usually throttle the smaller of the two flows.
A: A Coriolis meter will provide both mass and density and will require much less maintenance than turbine or PD. The accuracy will be better than 0.5% FS
A: Using a Coriolis meter is severe overkill. Since it is a chemical ejector, I am assuming that the pipe size is roughly 4 inches (100 mm nominal) or smaller. Your flow rate is reasonable for a 75-mm (3-inch) or 100-mm (4-inch) pipe. Your required accuracy of 0.5% of full scale means that the flowmeter is not required to be very accurate. Any flow element can be used here, from turbine to vortex to magnetic, since it is water. My favorite here, since you ask for high reliability and low maintenance, is a magnetic flowmeter.
A: I am a fan of magnetic flowmeters for this type of application, but a potential problem is the wide flow range needed. If you use a 75-mm (~3-in.) meter, the maximum velocity is about 9.5 mps (30 fps), which is on the high side for a magnetic flowmeter. This might cause liner wear or damage, but if it is infrequent, it might be okay.
If you use a 100-mm (4- in.) meter, the maximum velocity is about 5.5 mps (17 fps), which is much better, but the minimum velocity is only 17 mps (0.57 fps). Most magnetic flowmeters start losing accuracy under 1.0 to 1.5 fps. You should check the accuracy vs. velocity curves for the specific meter you intend to use to make sure it will meet your needs.
Curt Wendt, CAP