01 March 2004
Inferred mass flow finally gets respect
Coriolis is best of course, but enhanced old-school technology works.
By James Noel
With all the flowmeters available on the market and 50% of measurements being flow, it is no wonder that flowmeters are the most used, misused, and misapplied field device next to control valves.
From the early days of the Egyptians, engineers have been trying to measure flow while carefully balancing accuracy with performance. The market direction is toward Coriolis (mass) meters, but they are primarily for liquids; the compressible fluid processes have been left behind until now.
Additionally there are requirements for mass measurements where there is not a true mass measurement available. In this case there are many new options that use secondary process variables to calculate an inferred mass value. This article will look more closely at the inferential measurement techniques available today.
There is a strong push for inferential mass measurements through a combination of orifice plates and multivariable transmitters for mass values. The vortex continues to get better with age like a good wine, and in the wings we see a compensated multivariable version coming into play.
The magnetic flow tubes still get sliced and diced into market segments through various liner and transmitter combinations, so they are still a viable solution. The pitot tube came out of the closet to establish a relationship with the multivariable and is now respectable.
Now, compressible fluids such as gas and steam have a chance to grow up and get the respect they are due. Also liquid mass measurements are reasonable solutions with these supercharged and parameter compensated transmitters, as they can produce a reasonable mass value.
Only true mass measurement
First though, why are mass measurements valuable?
The reason process plants want a mass value is the same reason pilots of commercial airlines want to know exactly how much fuel they have while flying at 35,000 feet. We all realize that changes in pressure and temperature affect liquids and gases. Volume swings with changes in either condition. Thus, the ability to solve material balances within a plant or to control the exact amount of an expensive additive in a pharmaceutical plant or the butterfat content in a cheese plant can make the difference between profitability and liability—particularly in the case of prescription medicines. We calculate mass by multiplying the density by volume.
Having said all that, the reason Coriolis meters have taken off so well on liquids is that they are accurate, repeatable, and insensitive to viscosity. They also have no moving parts except for the vibration of the tube. When you look back on history, thanks goes to the French mathematician, Gustave-Gaspard Coriolis, whose theory in 1835 about the impact of inertial forces by a rotating reference (as in the case of Earth's rotation) is the basis for this measurement technology, which is the only true mass measurement available today.
At its resonant frequency
The Coriolis meter has been king of the hill for liquid mass flow readings for some time, and we expect to see it challenge the compressible fluid marketplace in the future. It is one of the two fastest growing meters and commands a large revenue stream. The Coriolis meter is best for liquids, though size is a limiting factor. It is insensitive to flow profiles, taking on almost any process with a wide rangeability in excess of 100:1. With no moving parts, maintenance is very low, but the initial cost is high. Total accuracy is better than 0.2% of rate and, let's face it, the meter is made for mass measurement and mass only.
The technology is based on a curved or straight tube vibrating at its resonant frequency. When the fluid moves through the tube it causes the tube to twist to the right, and this twisting motion is proportional to mass flow. Temperature measurements are made to compensate for metallurgical changes to the tube itself, but this technology—which you can easily demonstrate by turning on water in your home garden hose—works and works well. Another bonus is that the resonant frequency of the tube varies directly with fluid density. This measurement is independent from mass, and along with flow and temperature the transmitter provides all these independent measurements to the control system for other uses if needed (effectively replacing three transmitters). In summary, this is a pure bread mass device, but stay tuned to the marketplace and watch for the Coriolis meter entering the compressible fluid measurement arena. Then there will be a pure mass measurement for everything.
See the InTech magazine Tutorial/Control Fundamentals department's treatment of the Coriolis phenomenon and Coriolis technology at www.isa.org/intech/june2003/coriolis.
Adding the ANSI steam tables
Looking at the oldest volumetric technology in the pack that has moved into mass measurements, we find that the orifice plate has had a dose of new breath since the multivariable pressure transmitters came into the marketplace. The multivariable component added two key measurements that were missing totally or that happened using separate transmitters in the process piping somewhere near the primary flow measurement. The ability to compensate for pressure and/or temperature changes to the process has been the missing link to orifice plate credibility. These compensations were for compressible fluids where gaseous processes can swing the primary measurement in a nonlinear fashion from base conditions.
Say the volumetric flow of a steam flow is based on 150 pounds per square inch, gauge, and then the pressure dropped to 140 pounds per square inch, gauge. The resultant volume could be off by a factor of 10 pounds of steam, which at the hourly rate adds up fast. Now with the ability to compensate for changes in the base operating condition the result is not only correct, but you can also calculate the actual mass flow of the steam by adding the American National Standards Institute (ANSI) steam tables to the transmitter algorithm.
Instantaneously you have done a complete makeover of the orifice plate and put it in the class of the Coriolis meter for any line size. Plus you have improved the rangeability of the orifice plate from the typical 3:1 up to possibly 8:1. And don't forget the liquids it can measure even though neither the rangeability nor the accuracy is in the class of a Coriolis meter.
Averaging pitot tube cleaner
Another big winner in the primary measurement area is the averaging pitot tube, which has had no respect except as a last means of measuring large ducts, especially those that were not round. The aerodynamics of the averaging pitot tube have been greatly enhanced during the past few years in terms of keeping the back side of the bar clean from the typical process buildup. With the introduction of the multivariable pressure transmitter, it, like its older brother the orifice plate, has had new life breathed into it. Now those questionable measurements that we hoped would at least be repeatable are accurate for the first time, and they can convert to a mass flow reading if needed.
Because the multivariable transmitter can compensate for pressure and/or temperature, noncompressible fluids suffering from pressure and temperature upsets can receive proper and accurate analysis. The multivariable transmitter not only has the ANSI steam tables, but it also contains the AGA8 calculation for natural gas and the American Petroleum Institute standards for oils.
During transmitter configuration the compensation portion of the software allows you to use these reference standards in the algorithm to correct the volumetric reading, but you now have the choice on the averaging pitot tube to produce a mass value of that measurement as opposed to the volumetric reading.
The compensation is long overdue, because the rangeability of this primary measuring device has been on the short side. A 3:1 turndown has been the norm, just like the orifice plate, but with the addition of pressure and/or temperature compensation the rangeability has increased to 8:1. This makes this device a better alternative for gas, liquids, and steam applications, especially when the pressure changes in steam applications. Liquids are reasonable but again not in the class of a Coriolis meter.
Another head device winner in the volumetric arena is the V-cone that like the orifice plate and averaging pitot tube has come out smelling like the first rose of summer. This primary measuring device has mostly served in water and gas applications, providing a rugged differential pressure measurement feature by inserting a cone-style element in the flow stream.
With all the above features of the pressure transmitter's compensating software, it becomes an instant winner and therefore a more creditable measurement for mass applications. Now it is much more dangerous than it has been in the past. Its claim to fame is the ruggedness of the sensor, which is impervious to most suspended solids.
Traditional head devices
Recent abilities to capture pressure and/or temperature measurements from the process by the vortex flowmeter have made it a real solution for compressible fluid and liquid measurements. Since its introduction in the late 1970s the vortex meter has been a workhorse for volumetric measurements and a primary challenger to the traditional head devices, especially the orifice plate.
The vortex flowmeter has always had the larger rangeability and the linear output signal, but now with the addition of compensating measurements in a real-time fashion, the vortex meter has moved into another level of the playing field and become a tough challenger to the mass flow market.
The other advantage to the vortex flowmeter is low maintenance. With no moving parts it acts much like a turbine meter without the bearing problems. The vortex meter either works or it doesn't. The piezoelectric crystal is on or off, alive or dead. The resistance temperature detector (RTD) temperature sensor embeds downstream of the shedder bar so it won't disturb the flow measurement.
The vortex meter is a switch hitter able to measure liquids, gases, and steam. For steam applications the vortex only needs either a pressure or temperature compensating input, but for wet or super saturated steam it needs both. However, in the end it is still an inferred measurement for mass.
Another version of the vortex flowmeter is the swirl meter, which looks and acts like a vortex flowmeter but features the ability to handle upstream piping disturbances with ease. The meter has a set of blades at the inlet, which cause the fluid to swirl as it enters the body of the meter. The fluid accelerates through a reduction in the meter body creating a nozzle effect. A piezoelectric sensor located in the center of the meter measures the low-pressure zone, which has swirls that are proportional to flow.
Like a vortex meter the piezoelectric crystal measures the alternating low-pressure swirls. As the fluid is about to exit the body a deswirler helps eliminate the tangential velocity. Recently the swirl meter has incorporated a temperature sensor in the process stream, which makes it a contender in the mass measurement arena for liquids and gases.
The only limiting factors are the permanent head loss incurred by the process, the high cost of each meter, and the need for a larger line size than necessary to keep the flow laminar. With sizes up to 16 inches meters can cover the gamut of process applications and solve tight installation problems.
Another oldie but goodie is the tried and true magnetic flowmeter, which has been around since Faraday leaned over the bridge on the Thames River and found he could generate a current between two sensors. Companies introduced mag meters commercially in the early 1950s. Their foundation is Faraday's Law: flow of a conductive liquid through a magnetic field will cause a voltage signal to generate. The fluid must be naturally conductive, and the reading that eventually comes about is volumetric. Multiplying the volume by the fluid density yields a mass calculation.
This is as close as you can come to the Coriolis in terms of a true mass value. Magnetic flowmeters have a wide range of liners and electrodes available with their limitation being to conductive fluids only. There are both direct current and alternating current versions, as well as high-noise, electrode-less, and low-conductivity models. The jury on two-wire models is still out. In each case it is application and power dependent.
A secondary advantage of these new multivariable transmitters is the ability to utilize the compensating measurements elsewhere in the control system. If the user has a control system that can transmit digital measurements over Modbus, Foundation fieldbus, Profibus-PA, or HART then these measurements can work as primary measurements to other control functions. Many large control systems can use these measurements for other functions without installing a transmitter or sensor. An example is the use of temperature to monitor the process for alarming or changes to the control function, as opposed to just providing a measurement to the compensating algorithm.
Another use is providing the raw measurements to a remote control device that will do the compensating calculations, as opposed to these calculations taking place at the transmitter. Examples of this are drum density calculations for drum level in a power plant and a flow calculation in the remote terminal unit of a gas wellhead that feeds a pipeline. All in all, digital communications add value to these inferential mass measurement devices.
Inferred mass measurement devices fit markets where a Coriolis is not applicable. Although an inferred measurement may not be as accurate as the Coriolis measurement, it certainly has a legitimate place in measurement and control. Indeed, when it comes to large line sizes, rectangular ducts, or compressible fluids the inferential measurement is all that is available at this time. Let's face it; we have been dealing with inferred measurements for some time and surviving. Now, at least, we have some better tools in the tool shed.
Behind the byline
James Noel is a senior member of ISA. He has published 12 technical papers for ISA and the Technical Association for the Pulp and Paper Industry (TAPPI). He is the western region projects & strategic account manager, field instrumentation, for the Foxboro division of Invensys. Write him at firstname.lastname@example.org .
Coriolis goes low
The output—and accuracy—of most flowmeters can vary when any number of environmental and process variables, such as flow profile changes, temperature, pressure, viscosity, density, and specific heat, come into play. Calibrating and accounting for these variables in a device in a supplier's flow lab doesn't happen and is probably unrealistic.
Alternatively, in-process verification may improve the reliability factor, but the reference meter used to develop the in-process characteristics may suffer from similar effects.
It would be much better if the selected flowmeter technology was not susceptible to these effects at all. So what happens in a low-flow situation? There are very few low-flow metering solutions available on the market, especially in these very low liquid flow ranges (less than 1000 grams per hour) required for process research and development and pilot plant operation.
Coriolis mass flow technology, which has become popular over the past decade, has quite a few advantages: direct mass flow measurement and online direct density measurement based on fundamental properties, requiring no in-process recalibration or readjustment. Until recently, it was hardly possible to use a Coriolis mass flowmeter in real low-flow applications.
—Cor van Doorn, Brooks Instrument
Return to Previous Page