This article was printed in CONTROL's April 2009 edition.
By Walt Boyes
Ever since the invention in the 1790s of the Woltman-style mechanical turbine flowmeter, automation professionals have been looking for the one flowmeter that will work in every application. Unfortunately, there are 12 flow measurement technologies in common use for a very good reason. No single flow technology works well, or even acceptably, in all applications.
Of the more broadly based flow technologies, the one that works in the most applications, across most industries and with higher accuracy than even differential pressure is the electromagnetic flowmeter, or magmeter. According to Jesse Yoder at Flow Research, the total global market for flowmeters is roughly $4.7 billion, and magnetic flowmeters account for a little less than 20% of that total. There are a lot of magmeters shipped every year. Magmeters are used in every process industry vertical: water, wastewater, mining and minerals, utilities, food and pharmaceuticals. Magnetic flowmeters are designed for handling almost all water-based chemicals and slurries and are furnished with corrosion- and abrasion-resistant linings and even clean-in-place (CIP) designs for sanitary applications.
Magnetic flowmeters also are made in the widest size range of any flowmeter technology because they can be scaled up almost infinitely. The first use of the technology was in the huge sluices that drained the Zuider Zee in the Netherlands in the 1950s, and typically vendors supply a size range from ½ in. (12 mm) to 36 in. (914 mm), with several vendors supplying extended sizes up to 120 ins. (3048 mm). Several vendors sell sizes below ½ in. as well. How it is possible to scale up and down this broadly is directly related to the technology.
How a Magmeter Works
In 1831, Michael Faraday formulated the law of electromagnetic induction that bears his name. As used in an electromagnetic flowmeter, coils are placed parallel to flow and at right angles to a set of electrodes in the sides of the pipe, generating a standing magnetic field. The pipe itself must be non-magnetic and lined with non-magnetic material, such as plastic, rubber or Teflon. When the fluid (which must be conductive and free of voids) passes through the coils, a small voltage is induced on the electrodes, which is proportional to the deflection of the magnetic field. By Faraday’s Law, this deflection is the sum of all of the velocity vectors impinging on the magnetic field.
Figure 1. The velocity deflects the standing magnetic field and induces a voltage on the electrodes that is proportional to velocity.
Modern magmeters operate on a switched DC field principle to zero out noise that can be induced from RFI, EMI and electrical noise actually in the process fluid. They follow a regimen of turning the field off, measuring the voltage that is still induced on the electrodes, then turning the field back on and subtracting the off-state voltage from the on-state voltage. They do this several times a second, which reduces zero drift to almost nothing.
What this means for automation professionals is that the voltage induced on the electrodes is directly proportional to the average velocity in the pipe, and is therefore significantly more accurate than any other velocity-based measurement principle that only looks at a point or line velocity. In fact, the magnetic flowmeter is generally considered the most accurate wide-application flowmeter in current use.
Magmeter accuracy is remarkable, approaching the accuracy of positive displacement flowmeters. They’re often used for custody transfer when the flow is of relatively long duration. Typical accuracy of a magnetic flowmeter is 0.5% of measured value from 0.3 ft per sec to 33 ft. per sec (0.1 to 10 m per sec) velocity. Some vendors indicate even higher accuracies over portions of the flow range, up to 0.1% of indicated flow rate.
Where Magmeters Won’t Work
Magmeters have such a wide application that it’s easier to say where they will not work than to list all the applications in which they will.
They will not work when the pipe is not full (with the exception of several versions designed specifically for this application). If the pipe is not full, there will be significant error. One of the most common application failures of magnetic flowmeters is on a gravity-fed line discharging to atmosphere in a tank. Very often, at very low flows, the pipe is actually not full, and the flowmeter will read in error. If the pipe fill drops below the line of the electrodes, the meter will not read at all. Sometimes, applications like this are designed with a u-tube in the line, which is supposed to keep the pipe full at all times. And, sometimes this actually works.
They will not work when the pipe is full of entrained gas or air. This changes the computed volume of the pipe and changes the volumetric flow through the meter in an uncontrolled fashion that’s proportional to the amount of bubbles (or void fraction) in the pipe.
They will not work well where the flow starts and stops repeatedly because there is a lag between the time the flow starts and the correct velocity is read by the meter. This means that (again with the exception of some units that are specifically designed to be very fast) magnetic flowmeters don’t work well in short-duration batching operations.
They don’t read out in mass flow units, but when combined with an ancillary density measurement device (often, for larger diameter pipes a gamma nuclear densitometer), they can produce a high-precision mass flow measurement. This combination of devices is used regularly in any water-based fluid flow situation where the pipe size is larger than 12 in. (nominally 300 mm). This application is commonly found in the mining industry and in dredging applications in harbors and rivers around the world.
Most importantly, they will not work on non-conductive fluids or on gases at all. The minimum conductivity of a fluid is usually considered to be 5 μS (microSiemens) before a magnetic flowmeter will measure its velocity. In practice, it’s not wise to use a magmeter on a fluid whose average conductivity is this low.
Finally, magmeters (except again for specially-designed units) have trouble working on fluids with extremely high or highly variable conductivity. Saline brine and seawater, are examples of this kind of fluid.
There are some simple rules for using magmeters, which, if you follow them, will produce a satisfactory application.
Magnetic flowmeters need less straight run than most flowmeters, often as little as three diameters upstream of the electrode plane (the centerline of the meter body, usually), and no diameters of straight run downstream. However, there are circumstances in which a better choice is to go with as much straight run as you can get. For example, spiraling flow (swirl in the pipe) can propagate for hundreds of diameters after a three-dimensional turn in piping. Spiraling flow causes severe inaccuracy in a magmeter, sometimes as much as 40% of measured value.
“How a Magmeter Works
Figure 2. Some basic rules of thumb for using magmeters.
One of the ways to make sure you have a fully developed flow profile moving through the meter is to mount your magmeter so that the flow is through the meter in the vertical direction. This helps in cases of spiraling flow and also helps reduce air entrainment.
Although a magmeter will operate over the entire range from 0.3 fps to 33 fps (0.09 to 10 meters per second) velocity, it isn’t wise to install a magmeter that’s going to operate permanently at the lower end of that range. In applications where there are solids, this can cause buildup of solids inside the flow tube and sometimes on the electrodes themselves. If buildup occurs inside the flow tube, the calculated volume is now in error, and if buildup occurs on the electrodes, the insulating properties of the buildup can either reduce the voltage or break the circuit entirely. Either will cause inaccurate readings. It’s better to size the flowmeter for a normal flow that is about 60% of maximum for that pipe size, and if necessary, install a properly designed meter run. Fortunately, for a magmeter, that meter run doesn’t need to be as long as it does with some technologies for measuring flow.
Remember that the pipe section of the magmeter needs to be non-conductive for the circuit to work. The electronics that process the induced voltage, however, are susceptible to interference if they’re floating above ground. Magmeter vendors all have grounding procedures, which you ignore at your peril.
Temperature and Pressure
Magnetic flowmeters are designed to work at moderate temperatures and pressures and should not be stressed above or below those specifications. Magnetic flowmeters should not be operated where a vacuum can be pulled inside the flow tube unless specifically designed for that service. This is so especially when there is a pressed-in polyurethane or Teflon lining, because the vacuum can pull the lining right out of the meter, causing potential hazard, as well as inaccuracy in reading. Both Teflon and polyurethane, which are the most common magnetic flowmeter liners, are de-rated for pressure at the upper end of their temperature range and will deform if overheated.
Magnetic flowmeters have become one of the most widely used flow technologies in the 50 years since their first introduction. They’re simple, easy to maintain, and because they have no moving parts, able to operate for years without maintenance.
Walt Boyes is Control’s editor in chief.