Adaptive sensor technology in the chemical industry

A recent installation at BYK-Chemie in Wesel demonstrates how a mass flowmeter's design impacts the accuracy of the measurement for products with different viscosities.

The case is a good demonstration of adaptive sensor technology paving the way for successful mass flow measurement in the chemical industry.

Mass flow measurement at BYK-Chemie

BYK-Chemie is a member of the Altana Group and a leading supplier of coating and plastics additives worldwide. These are used in minute quantities and improve the qualitative properties of finished products. This specialty chemical company is one of the most innovative in the sector and highly values research and development and customer orientation. This is why BYK-Chemie operates application laboratories around the world, enabling the company to test new processes and products.

Originally, BYK-Chemie used straight-tube mass flowmeters in dosing various chemicals with differing viscosities. Up to eight chemicals needed to be added sequentially to a container during the production process for coating additives.

After several faulty batches, BYK-Chemie did some troubleshooting to determine what was causing the problems, particularly since the high price tag for chemicals was quickly driving the total cost to an unacceptable level.

During the investigation, measurements taken using precision scales revealed deviations of various magnitudes for different media. The manufacturer stated that these were clearly viscosity related.

The supplier then recommended that BYK-Chemie replace the straight-tube instruments with twin-tube meters. If this recommendation had been followed, the dramatic increase in dosing times would have meant additional shifts for the company. Furthermore, one of the stated objectives was that there should be no appreciable drop in pressure when dosing.

This is not possible with twin-tube meters. BYK-Chemie therefore rejected the recommendation. As an alternative, the supplier recommended a zero-point adjustment for every product change. This also would have led to longer dosing times. Zero-point comparison takes several seconds and is very complicated to do automatically. BYK-Chemie therefore decided to test and directly compare Krohne's Optimass with the original mass flowmeter.

Fine dosing: the comparison test

It is standard practice to thoroughly test every flowmeter against all specifications before shipment. This characterisation of the instrument goes far beyond normal calibration. All of the meter's calibration constants are set in the meter and its metering behaviour, including zero-point stability, linearity, mass and density measurement accuracy and repeatability, is tested at a minimum of three different product temperatures and at various flow rates.

All test results are documented and are available at any time to users. Established standards recommend adjusting the zero-point on a Coriolis mass flowmeter on site. When it comes to metering highly viscous media, the readings usually fall within the low end of the measurement range. For these types of applications, reliable zero-point stability plays an extremely important role in achieving very precise measurements.

BYK-Chemie decided to use Krohne's mass flowmeter straight from the factory and not to do the zero-point adjustment on site when it was tested against the Coriolis meter they were presently using. An Optimass with a nominal diameter of 40 mm was installed vertically in series and flange-to-flange with the existing unit during the test run. BYK-Chemie then measured the mass flow rate for three products with different viscosities on both units.

The company did several test runs for each product. During the test, the initial flow rate of up to 25,000 kg/h was reduced to a maximum of 1000 kg/h for fine dosing. Both units met the user's accuracy specifications and had a stable zero point when comparative measurements were made on Xylene, a low viscosity product (see Figure 1). When the product was changed to Bupol, which has a viscosity of 200 mPa, major shifts in the zero-point occurred with the first meter.

Even after stopping the flow, the counter readings continued to change. These were corrected when the measured values were recorded. Every measurement taken during fine dosing had over 10% error.

The last product tested was Polyglycol, which has a high viscosity of 400 mPa and a low flow rate. The measurement error increased even more, reaching almost 15%, while the Optimass readings remained reliable and accurate.

After completing the comparison test, BYK-Chemie has tested T08, T10, T25 and T40 units in various applications. Exact and reliable measurement is essential to BYKChemie's process, particularly for mass flowmetering at the end of the distribution heads from which various products are delivered to the mixing vessel.

The results of the mass measurement are fed to a process control system (DCS) that controls the dosing valves for about 35 production units. Up to 20 different substances required to produce BYK products can be dosed from each distribution head.

At this point the question "Why is the replacement straight-tube meter so much better than the unit used up until now?" deserves to be answered. Part of the answer lies in the differences in the design. For example, the first meter installed is also supposed to be able to detect the viscosity of a particular product. The application we described here confirms this statement. However, the actual mass flow readings become inaccurate with changing viscosities. In other words, the meter readings are dependent on viscosity.

The second part of the answer is related to the design and the electronics used. By employing adaptive sensor technology in the design, Krohne has succeeded in making its Coriolis straight-tube meter independent of process conditions.

In addition, compared to conventional mass flowmeters, the electronics used in the Optimass have significantly improved the measuring behaviour at the low end of the metering range. This enables users to very precisely measure even the low flow rates that occur with highly viscous media, without having to reduce the pipe diameter, which in turn would lead to a higher pressure loss. Since the units are able to measure low flow rates over a wide range of nominal diameters, overdosing does not occur and meter readings remain accurate.

The Optimass straight-tube design

Krohne has always focused on the straight-tube design, since the advantages over mass flowmeters with bent or double tubes are clear. Curved or o-shaped tubes and twin-tube designs with flow dividers have proved unreliable in many applications.

For example, it is very difficult to meet certain chemical and pharmaceutical industry requirements when using meters with flow dividers and whose flow paths are not straight because the specifications call for no dead space or crevices. Some media, such as viscous, non-Newtonian, shear-sensitive fluids or solids-bearing fluids cause high pressure drops in twin-tube meters. Abrasive media can wear away flow dividers and tube bends, and metered media containing fibres, such as palm oil or cellulose, can lodge in the flow dividers and cause blockages.

The most important benefits of straight-tube flowmeters include:

• low pressure drop
• less wear due to abrasion
• also suitable for high flow rates
• wide measuring range
• no risk of blockage

In addition, straight-tube flowmeters are easy to clean and self-draining. The favourable relationship between the short measuring tube and large inner diameter makes the unit suitable for measuring even highly viscous media.

Each individual application will determine whether users will benefit from each and every one of the advantages of Coriolis straight-tube flowmeters as described above. One of the top priorities in the development of the instrument was to design a meter with high accuracy over a wide measuring range while maintaining good zero-point stability. Various factors can influence the measuring accuracy and zero-point stability of Coriolis flowmeters, including the gas content of fluids, the homogeneity of solid-bearing fluids and also viscosity. The BYK-Chemie experience demonstrates how large an impact viscosity can have on a mass flowmeter's readings if it is not designed properly.

Adaptive sensor technology (AST)

Since its launch, the technology has been used successfully in many applications. AST represents a quantum leap in the advancement of Coriolis technology that has enabled Krohne to make the Optimass immune to process vibrations and external process conditions and to deliver precise readings.

Conventional straight-tube mass flowmeters often use counterbalancing absorber masses to mechanically decouple the measuring system from the process. These consist of a spring-loaded mass installed on the inner cylinder. On the Optimass, the counterbalancing absorber mass and spring have been replaced by an integral mass and spring system that uses the dynamic elements of the measuring system. For conventional straight-tube meters, the absorber masses and the spring stiffness are constant, but the density of the product and the vibrating frequency vary.

As a result, the compensation provided by the counter-balancing absorber mass is only effective over a very limited density range. On the other hand, the AST system consists of an inner cylinder where the ends of the connecting tubes act as a spring for the absorber mass system. The main benefit is that this 'spring' (ie, the measuring tube-ends) is self-compensating. As the density of the fluid changes, the dynamic behaviour of the tube-ends (or spring) also changes in relation to the inner cylinder mass. This optimises the measuring performance and results in a fully balanced system under all operating conditions. Since all the vibration energy remains inside the meter, the zero point remains stable and the measuring accuracy is improved.

The Coriolis measuring principle within a straight-tube Coriolis mass flowmeter is able to derive mass flow from the tube deformation that is caused by the inertia of the mass flowing through it. The density of the product can be determined by comparing the tube's vibration frequency when the product flows through to its natural resonant frequency. The Optimass sensor consists of a straight measuring tube that is caused to vibrate by an excitation coil mounted at the mid-point of the tube. The exciter control circuit ensures that the measuring tube continuously vibrates at resonant frequency. Two sensors mounted at an equal distance from either side of the exciter coil are used to detect the Coriolis effect. With no flow, the two sensors deliver the same sine wave signal, which corresponds to the tube's natural vibration frequency. As soon as product flows through the tube, Coriolis forces impact the flowing mass particles, causing the measuring tube to deform and thereby create a phase shift between the two sensor signals. The sensors measure the phase shift between the two sine wave vibrations. The phase shift is directly proportional to the mass flow rate.

Mass flowmeters are being applied in ever-increasing quantities since they are able to precisely measure liquid and gaseous products thanks to the Coriolis principle. Coriolis instrumentation is the number two flow measurement technology, right behind electromagnetic flowmeters. It and ultrasonic flowmeters are exhibiting the highest growth rates.