Company Announcement -Dr. Nicolaus Mathies, Krohne Messtechnik GmbH, Duisburg

Measurement of flowing liquid in closed conduits is present in many different applications in our daily life, e.g. water or heat metering, gas metering, filling stations, in the chemical industry, the food industry, for batching or for process regulation. Usually liquid flowmeters are used to fulfil most of these requirements. The demands of the application on flow meters are high. We have a general expectation that meters “show the correct value”, which is necessary for correct billing. Without external help it is normally impossible to check whether a flowmeter shows the correct values or not. Behind the accuracy of flow meters is a relatively complex system of calculations, standards, normative documents and special calibration facilities with special knowledge of the operators. This article describes the requirements on liquid flow meters with regard to the measurement accuracy.

Introduction:
The transfer of measurable goods is regulated metrologicaly in the area of custody transfer measurement by the European Measuring Instruments Directive 2004/22/EG [1] the German Mess- und Eichgesetz [2] and the Eichordnung [3]. The requirements for suitability, usability and application area are covered under general definitions. Special requirements for flow meters are defined depending on the different applications. Maximum permissible errors (MPE) depending on the measurand (m³, kg or kWh) define the “accuracy requirements” for the meters, which have to be fulfilled by every single meter. Many technical official standards and documents are available which the basis for comparable and repeatable measurements is.

In industrial metrology, many referenced guidelines for process applications exist to ensure correct installation of flow meters, where the application can influence the meter measurement. To achieve reproducible results and meters which can be replaced one for another; the flow meters have to be calibrated with the same reference conditions, which lead to traceable and comparable measurements.

Traceability of measurements
The transfer of measurable goods generally requires a consistent measurement of the units which are used like kilogram or meter. The question about a standard definition of the measurands was discussed by the ancient Egyptians and Romans. The goal of the discussions was to achieve equal, comparable and reproducible values of a measurand so that a fair trade of the goods could be realized. Definitions of reference lengths and references weights can be found in a lot of different areas of measurement. For a long time in Europe these definitions were established by the local kings. The result was a huge number of different standards for the kilogram, for example, and difficulties in conversion of these standards. In 1875 with the Meter convention 17 different European countries signed a contract to use the same reference kilogram and the same reference meter and to establish an institution to keep these standards. For the first time this allowed traceability on the same standards. Up to the present 51 states have signed this contract [4] and there are two associated states. In addition to meter and kilogram, five new measurands have been added. Since the 13th general conference of mass and meter in 1960 the system is called SI-System, the international unit system. The SI-units today are kilogram, meter, second, Kelvin, ampere, mole and candela. The transfer and traceability of the SI-units has to be done by the national metrology institute of the different countries. In Germany this is done by the Physikalisch-Technische Bundesanstalt. The technical definitions of the SI-units can be found at the homepage of the PTB [5]. Next to the SI-units several additional derived units exist, for example typical flow units as l/h, l/s, kg/s or kg/h.

Reference methods
Liquid flow measurement is generally used for example for the measurement of cold potable water, heat metering applications, gas measurement or fuel dispensers, truck, rail, or ship loading applications. In addition, industrial processes are sometimes coupled to flow measurement. For the calibration of all these flow meters, different reference methods can be used, depending on the required measurement uncertainty of the meter under test and the application.

Reference meter:
The meter under test is calibrated with another flowmeter of the same or different construction. Usually more accurate master reference flowmeters which have been specially selected are used. The metrological behaviour (measurement uncertainty, repeatability and reproducibility) of those meters is well known and the meter has a long calibration history. If liquid flowmeters are to be calibrated, usually calibration rigs with a circulating flow are used. The medium is taken out of a collection tank and pumped into the measuring section, where the meter under test is mounted. Within the defined calibration time the meter under test is compared with the reference meter. Sometimes the error curve of the reference meter is corrected during the calibration of the meter under test.

Of course the minimum measurement uncertainty which can be achieved is at a medium level, because a reference flowmeter must be calibrated elsewhere with another reference standard with a smaller measurement uncertainty. Depending on the physical principle of the meter under test, the reference meter method is used for many industrial calibration processes.

Fig 1 shows a liquid flow calibration rig with several Electromagnetic reference meters.

Fig 1: Liquid flow calibration rig with Electromagnetic flowmeters as references

Gravimetric weighing system
With more stringent requirements of the measurement uncertainty for flowmeter calibration, liquid flow calibration rigs are used as a reference [6]. The flowing liquid is measured by a weighing scale (mass), a temperature measurement (density) and a time measurement see fig 2. The medium is taken from a collection tank and directly (or with a constant header tank, not shown in fig 2) pumped into the measuring section. The meter under test is mounted in the measuring section. The medium passes through the meter under test and flows through a reference and flow control valves into the weighing tank. The calibration method shown is the standing start stop method, where the medium is directly pumped into the weighing tank from the start of the pumps. If the flying start stop method is used, the medium is diverted with a liquid flow diverter during the measuring time into the weighing tank. In the meantime the liquid is pumped back into the collection tank [7, 8]. The standing start stop method avoids the influence of the liquid flow diverter, but of course the error curve influences the result of measurement uncertainty of the calibration.

fig 2: gravimetric weighing system

Volumetric calibration system:
If volumetric measurands have to be calibrated, then volumetric references are used for the calibration. The volume of a flowing liquid is compared with a tank or prover system or any kind of reference volume with a small measurement uncertainty. As mentioned for the gravimetric weighing systems, the volumetric calibration systems can be used in the standing start stop method if a reference volume is used with or without a liquid flow diverter. For example, a concrete volumetric reference tank is used in conjunction with level measurement at the local Weights and Measures authority in Düsseldorf, Germany for the calibration of cold water meters. Compact systems with a smaller maximum flowrate than the reference tank systems are realized today as prover systems. Here the reference volume is defined by a cylinder. A piston displaces the medium and brings the medium through the meter under test. These systems are limited for the maximum flowrate and maximum velocity through the meter under test.

Fig 3: volumetric reference system

Validation of measurements
The validation of measurands for liquid flow measurement is a more complex thing compared to the other SI-units. A direct comparison as it is done for mass units is not possible because of the complexity of the liquid flow calibration systems. For a description of the measurement accuracy for the calibration systems, measurement uncertainty calculations are used [7] and afterwards validated with key comparisons between the calibration rigs. This is usually done with reference flow meters, which have been tested and characterized for a long time so that the measurement history of the instrument is well known. These liquid flow meters are called comparison standards, because their flow scale is used for a comparison of the flow scale of different liquid flow calibration rigs. Key comparisons can be done in different ways, see fig. 4.

Fig 4: Different structures of key comparisons

The structure of a key comparison can be realized as a circle or a star, as shown in fig 4. Comparisons which are organized in a circle rotate the reference flowmeter after it has been measured in the pilot laboratory. At the end of the circle it is measured in the pilot laboratory again. Measurement errors which occur when the reference meter is measured at the different laboratories can’t be detected during the comparison so that establishing the time of the error is difficult. A comparison which is organized in a star structure improves detection of measurement errors. Here the reference meter is measured at the beginning in the pilot laboratory and afterwards in another laboratory which sends the meter back to the pilot laboratory. In this case measurement errors can be detected quickly. In practice combinations of both structures are used. The District Heating Working Group (AGFW) key comparison uses a combination of the circle and star structure for a comparison of calibration rig for the flow sensor of heat meters [9]. Before a comparison takes place, the goal of the comparison, the measurement uncertainties of the calibration rigs, flow rates, pressure, temperature of the medium and the measured quantity have to be precisely defined. The investigation of a standard calibration including calibration procedure and operators has been done in [10].

Conclusion
The measurement of flowing liquids in closed conduits is present in a lot of different applications of our daily life like water, gas or heat metering or batch applications. Flowmeters are often used to fulfil these requirements. Normally we accept the quantity of the measured medium which is displayed by the flow meter without being able to check technical details of the meter. For this, a wide range of technical descriptions, approvals, standards etc are available. Liquid flow meters are calibrated depending on the requirements for the measurement accuracy with different reference systems like reference flow meters, gravimetric or volumetric calibration systems.

Therefore a lot of different influences on the reference systems have to be taken into account. These are described in a measurement uncertainty calculation of the system. The validation of these calculations can be done with key comparisons, which considers the expected measurement uncertainty of the comparison, the rigs, and different reference flowmeters. The result of the sum of requirements for the comparison provides a general way to compare liquid flow measurements at different locations with a traceable accuracy and acceptance of the meter measurement trueness at the end customer.

literature

• Measuring instruments directive (MID), 2004/22/EG
• Gesetz über das Mess- und Eichwesen (Eichgesetz) vom 23. März 1992 (BGBl. I S. 711), zuletzt geändert durch: Art. 2 G vom 3. Juli 2008
• Eichordnung vom 12. August 1988 (BGBl. I S. 1657), Zuletzt geändert durch Art. 3 § 14 G v. 3.12.2007 I 2930
• BIPM, Meterkonvention, www.bipm.org
• Physikalisch-Technische Bundesanstalt, www.ptb.de
•  Mathies, N.; Adunka, F.; Schupp, R.; „Volumenstrommessung mit gravimetrischen Normalmessanlagen“ tm, Technisches Messen, Ausgabe 11-2008, Oldenbourg Wissenschaftsverlag, München Oldenbourg Wissenschaftsverlag, München
• Mathies, N.; Engel,R.; Baade,H.-J.: Modellbasierte und messtechnische Validierung von Umschalteinrichtungen in Durchflusskalibriereinrichtungen für Flüssigkeiten, tm Technisches Messen, Ausgabe 5-2010
• Mathies, N.: „Messunsicherheit einer gravimetrischen Kalt- und Warmwasser Normalmessanlage für große Volumenströme“, Dissertation, Mensch- und Buch Verlag, Berlin, 2005
• Leitgen, G.; Espig, F.: „AGFW-Ringversuch für Prüfstellen für Wärme-Messgeräte“ EuroHeat & Power, Heft 1-2/2010, VWEW-Verlag, Frankfurt
• Mathies, N.: “Präzisions-Durchflussmessung mit Normalmessanlagen” ATP, Automatisierungstechnische Praxis, Ausgabe 7-8/2010, S. 42-49, Oldenbourg Industrieverlag, München

Dr. Nicolaus Mathies
KROHNE Messtechnik GmbH,
Ludwig-Krohne-Straße 5, 47058 Duisburg
Tel.: 0203 / 301 4182; Fax.: 0203 / 301 10182
E-Mail: n.mathies@krohne.com

Contact: KROHNE South Africa : www.za.krohne.com  011 314 1391     salesza@krohne.com