Applying Coriolis technology to high-pressure applications

The first commercially available Coriolis flowmeters based on the observed effects of mass flowing through vibrating tube systems appeared in the late 1970s and early 1980s. Any mass flowing through such tubes causes dampening and distortion effects in the tubing system, which correlates to the actual flow in the piping. Today, many Coriolis mass flowmeters in use employ this technology, with about a dozen different manufacturers producing them. All of these meters are built on a vibrating piping system with the inertia of the mass of material flowing through creating very small but measurable deflections of the tubing.

The name for these meters comes from the force responsible for the deflections - the Coriolis force. Compared to other technologies, which mostly determine flow velocity, Coriolis mass flowmeters offer direct mass flow measurement and, unlike velocity measurement techniques, changes in density, viscosity and flow profile do not, in general, play a significant role when measuring flow with a Coriolis meter.

These tremendous advantages, along with a non-invasive nature, lack of moving parts (which together equate to a high turndown ratio and minimal maintenance) and high intrinsic accuracy (typically 0.15-0.5%) have made Coriolis meters very desirable as measuring elements. Coriolis meters are traditionally higher priced than many other metering technologies, but over the last few years prices have dropped due to designs oriented more to bulk manufacturing techniques and resulting economies of scale. Manufacturers have responded to the requirements of specific industries by offering dedicated solutions to applications such as natural gas dispensing, and this has led to further adoption of Coriolis meters in wider market areas.

Excitation force applies

Coriolis mass flowmeters have different designs with regard to shaping the tube system for measuring the flow. The U shape was initially one of the most popular tube geometry designs, and it serves to illustrate the theory of Coriolis flowmeter functionality.

The principle of operation is as follows. Application of an excitation force to the U-shaped tubes causes them to oscillate backwards and forwards, while flow enters the tubes at one side of the meter, travels through the tubes and exits at the other side.

The oscillation of the tubes is orthogonal to the material flowing within them. As material passes through the tubes, the flowing mass accelerates in the direction of the oscillation. Due to its inertia, the tubing sees a force - the Coriolis force - that adds to the deflection of the tube around the oscillation axis. The tube form takes on a double-bended or ‘S’ shape. This additional bending registers as a phase shift and is directly proportional to the mass passing through the tubes.

The additional bending, which is most pronounced in the middle of the U shape, is a direct result of Coriolis force and relates only to the mass moving through the meter. The more mass flow, the stronger the Coriolis force and the more pronounced the bending. The range of bending is very small, typically in the range of one ten-thousandth of a millimetre up to a few tenths of a millimetre, depending on design. For accurate, stable mass flow measurement, a good signal-to-noise ratio is necessary - a high signal means significant bending and deformation of the tubes while low noise requires external factors like vibrations to contribute minimally to the primary measured deformation. Both factors strongly relate to the design of a Coriolis flowmeter.

The design of a meter influences the strength of the induced Coriolis force by its geometric layout and the amplitude and frequency of oscillation. To achieve larger tube deformation and therefore more precision in measurement, fast and energetic oscillation and a long ‘arm’ (the distance from the upper U shape to the oscillation axis) is desirable. The momentum responsible for the deformation is proportional to these parameters. By contrast, working against the deformation by the Coriolis force is the elastic modulus of the tubes - here called the spring constant. The higher the spring constant, the less deformation that registers at a particular Coriolis force. Considering this, the design goals for a Coriolis flowmeter should not only consider the creation of a large Coriolis force, but also the spring constant of the tubes themselves, as this can work against large deformations or deflections.

Considering these interdependencies, at first glance the ideal Coriolis mass flowmeter should have the following features:

• Long, easily bendable tubing systems
• Large distance between oscillation axis and excitation point
• Thin wall tubing to keep the spring constant small
• High energy excitation to create large oscillation amplitudes

Unfortunately, some practical aspects of meter design temper the design goals for the ideal meter:

• Long piping systems may create unacceptable pressure drop
• High-pressure, abrasive or corrosive media require thick wall tubing
• High energy input may conflict with safety requirements in hazardous areas
• Excessive excitation may lead to fatigue failure of the tubing

The above, while strongly simplified, shows there is no one ideal Coriolis mass flowmeter design. Regardless of where design emphasis lies, there will always be trade-offs made for applicability and practical suitability against performance.

It is worth noting the performance of a Coriolis flowmeter can suffer from external noise, and installation quality remains a key factor in Coriolis flowmeter performance. While installing the meter in a vibration-free environment can minimise external noise issues, the design of the meter itself must be optimal to prevent problems. Here the criterion is to keep the system oscillation stable and difficult to disturb from the outside. It is important that balanced tubing is part of the hardware so the swinging is very stable and possibly self-sustaining - like a tuning fork - and the actual point of deformation measurement is well decoupled from any process connection stress and influence.

Over time, the various Coriolis meter designs have improved and are now nearly all suitable for the ‘run-of-the-mill’ applications found in everyday processing. However, where meters are exposed to extreme applications and conditions, the differences and limitations of various designs can become noticeable.

One design that demonstrates advantages over others in extreme applications is the omega tube torsion design. Omega tube meter features The omega tube torsion Coriolis mass flowmeter is unique in that it is equipped with torsion rods and crossbars as part of its mechanism. This design has proved to be universal, serving ‘standard’ mass flow applications and, more importantly, extreme applications with high flow rates, pressures or temperatures: Rates up to 1500 m/h Line sizes up to 30 cm Pressures up to 900 bar Temperatures from –250 to 400°C

Figure 1: Design features of an omega tube meter.

The omega tube design is different from traditional Coriolis meters. The active part of most Coriolis mass flowmeters consists of its oscillating tubing, while the omega tube system consists of three distinctly different mechanical elements, each dedicated to an essential function of the flowmeter:

• Half-circle measuring tube - The Coriolis force deforms this part, and it is therefore the active measurement element in the meter. Installed here are deformation sensors in the form of pickup coils and magnets.
• Torsion bar oscillation system - This system consists of two torsion rods and two crossbars providing the base oscillation system. It would oscillate even without tubing attached.
• Process feed tubes - This section is below the mass bars and sees almost no bending from the meter oscillation; it sees only a low-stress torsion moment.

The separation of the functional elements gives the opportunity to optimise each element separately according to its function and mitigate the trade-offs between the ideal design in the sense of creating strong measurable signal, and application requirements, such as heavier wall thickness and wetted material selection. The active measurement element is fixed left and right where it passes through the crossbar. Since the section is an exact semicircle, pressure changes - even the highest ones - do not alter its shape. Other designs have a natural tendency to slightly lose the shape of their original form when pressurised and this affects accuracy. This means an omega tube meter calibrated on water at low pressure, for example, can serve to measure compressed natural gas in the field at a pressure between 200 and 300 bar without affecting precision or zero. External noise effects remain at a minimum because the measurement section is isolated from the majority of such noise by the mass of the crossbar and, apart from the pick-up devices that measure the deflection induced by the Coriolis force, the section is free of additional elements that could resonate and disturb the measurement.

Figure 2: Three-inch omega design meter with ANSI Class 2500 flanges.

The real key to the omega tube flowmeter’s suitability to difficult applications is the oscillation system. The system’s characteristic components are the torsion rods and the crossbars. Each torsion rod with its crossbar represents an oscillation system on its own. It is like a tuning fork that works independently, even without attached tubing. Excitation energy for oscillation is injected via coils sitting on the crossbars themselves. The torsion rod serves by storing this injected energy, smoothly delivering it into the oscillation movement as the tubes swing back and forth. The use of crossbars in conjunction with torsion rods creates very energetic and stable oscillation with very little energy input and, once oscillating at harmonic frequency, the meter is very mechanically tolerant of disturbing and dampening effects because of the mass of the crossbars. A standalone tubing system does not build up and keep this amount of oscillation energy. The energy requirement of an omega tube meter to sustain oscillation is in fact so low that even full 15 cm piping with a wall thickness of over 5 mm can still be rated intrinsically safe for zone 0. Furthermore, the large amplitude of oscillation generated is not critical in terms of mechanical stress to the tubing system. Unlike in a classical U-shape configuration, a large movement in the active section only results in small non-critical torsional movement in the process feed tube section and not in an overbending of a tube. This feature makes it possible to use heavy wall thickness tubing while still allowing the generation of large amplitude oscillations.

For most Coriolis mass flowmeters, the tubing actually represents the spring constant, and hence changes in the tubing directly influence the operation. Tubing in the omega tube design plays a secondary role with regard to the spring constant of the meter, and changes in the tubing type can easily be accommodated without major re-engineering, allowing the use of very different materials from standard stainless steel to exotic (with very different elasticity modulus) materials like tantalum.

Extreme diameter-to-wall thickness ratios are not a problem either. For instance, tubes with an outside diameter of 114 mm and a wall thickness of 11 mm work without degrading performance. The ability to design meters with standard wall thickness tubes rather than thin wall tubes also means much higher pressure applications can be metered - ANSI Class 900, 1500, and 2500 flanges are not uncommon on omega tube meters - along with the welcome advantage of not having to fit the necessary secondary pressure housings needed to provide containment in the event of a tube failure. The use of thick wall tube and pipe within the omega tube meter translates into high confidence in safety when installed on site.

*John Daly has more than 30 years of experience in instrumentation and control. He works as the lead product specialist with GE Sensing for the Rheonik Coriolis Flow Meter brand.

Resources

• The two phases of Coriolis flow www.isa.org/link/Jan2009Coriolis
• Coriolis: Twist and Shout www.isa.org/link/200306Coriolis
• Flow Research, Inc. www.flowresearch.com

This article was first published in the January/February issue of InTech magazine, © 2010 ISA, and was reprinted by permission.