Inventory and basic evaluation of fi eld
calibration methods for volume and
fl ow meters
Nordtest 1590-02
SP Measurement Technology SP REPORT 2004:24
SP Swedish National T
Inventory and basic evaluation of fi eld
calibration methods for volume and
fl ow meters
Abstract
Inventory and basic evaluation of field calibration methods for volume and
flow meters
Volume measurement of liquids is of great importance in many industrial processes. Nordic industries such as process industry, petroleum refineries, district heating, water supply and pulp-industry are examples of industries in need of improved volume measurements with traceability. Some of the meters in this area are either to large to calibrate in a laboratory, or runs in a process liquid which is not "compatible" with existing calibration liquids. An on site calibration would be preferred for these meters, but today the possibilities for this are very limited. Further more existing methods have limitations that often are unknown to the users or at least have not been looked at thoroughly from a measurement uncertainty point of view.
The scope for this project is to evaluate methods and produce a preliminary guideline for finding the “right” method and to know the limitations of commonly available methods.
SP Sveriges Provnings- och SP Swedish National Testing and Forskningsinstitut Research Institute
SP Rapport 2004:24 SP Report 2004:24 ISBN 91-7848-998-9 ISSN 0284-5172 Borås Postal address: Box 857
SE-501 15 BORÅS, Sweden
Telephone: +46 33 16 50 00 Telefax: +46 33 13 55 02
Contents
1 Introduction 5
1.1 Background 5
1.2 Problem description and definition 5
1.3 Realization 5
1.4 Exclusions 6
2 Market survey 6
3 Workshop with instrument suppliers 8
4 Existing Calibration methods 8
4.1 Tracer methods 8
4.2 Clamp-on ultrasonic flow meters 9
4.3 Calibration using “built in” equipment 10
4.4 Standardisation in the field of meter calibration 10
5 Practical problems 11
5.1 Access to meters 11
5.2 Signals 11
6 Evaluation of methods 13
6.1 Test 1 – Magnetic inductive flow meter as master 13
6.2 Test 2 – Ultrasonic clamp-on flowmeter at different pipe materials 14 6.3 Test 3 – Ultrasonic clamp-on flowmeter at different liquid temperature 15 6.4 Test 4 – Ultrasonic clamp-on flowmeter at different pipe sizes 16 6.5 Test 5 – Ultrasonic clamp-on flowmeter at different positions 17 6.6 Test 6 – Using a ultrasonic clamp-on flowmeter as master for calibration 18
6.7 Test 7 – Using the tracer method for calibration 21
6.8 Evaluation at other institutes 25
7 Preliminary guide for the selection of methods 27
7.1 Volume standards 28
7.2 Master meters 28
7.3 Clamp-on ultrasonic meter 30
7.4 Transit time tracer method 30
7.5 Other methods 30
8 In the future… 30
8.1 Ongoing research at institutes 30
8.2 Research at manufacturers 31
8.3 Combination of methods and principles 31
8.4 User Training 32
9 Uncertainty and traceability 32
10 Conclusions 32
Summary
A market survey for methods and equipment suitable for field calibration of flow meters has been done that included a number of sources such as databases, a questionnaire to suppliers etc. The result was that only three general methods for field calibration of flow meters were found. These are reference meter placed in-line, clamp-on ultrasonic meters and tracer injection methods.
Technical meetings have been held with several companies that are active in this field. A workshop was held at SP in Borås 2003-10-09.
Two-degree works at the University College of Borås and Chalmers Lindholmen in Gothenburg have been initiated, performed and presented. A search for international research work has also been done. Practical field tests were planned within the project. These tests did however met with difficulties, as the planned industrial sites (petroleum, district heating and pulp) cancelled out of practical reasons. It has thus not been possible to find suitable locations with possibilities for multiple measurements. Instead simulated tests have been made in laboratory environment. As it is not possible to simulate industrial realities like large diameter pipe works or liquids with high concentration of solids, these laboratory tests must however be considered as relatively “easy” regarding environment and flow conditions.
The conclusion of this project is that limitations of the existing, practically available, calibration methods have been confirmed.
We are grateful for the cooperation from Risto Kuopamääki from Indmeas OY and Per Fredriksson from G E Panametrics and their colleagues. We also send our thanks to Davor Popovic and Anders Ahlsten at Borås Högskola as well as Mikael Johansson and Samuel Ekström at Chalmers
Lindholmen (including teachers and supervisors) for their contribution.
We also like to thank the participates from Fortum (Eva Katrin Lindman), Södra Cell (Sigurd Björkman) and Preem (Olof Olsson) incl. all their colleagues for the support. Even if we failed in finding suitable locations for field measurements interesting discussions have taken place and possibilities are high for future cooperation in other projects.
1 Introduction
1.1 Background
Increased prices for energy and raw material, requirements for stricter quality control and more efficient control of wastewater flows, are some factors that have resulted in an increasing need for the calibration of volume- and flow meters. Historically, calibration has been done mainly for custody transfer meters and relatively small meters. Today there is need also for calibration of other types and sizes of meters.
There are several thousands of patents in the flow measurement area and the number of available flow meter principles is very large. All these meters have different technical solutions with advantages in some applications and disadvantages in others. This can make the selection of a suitable flow meter to a specific application quite difficult and the same apply when selecting a suitable calibration method. Some meters can easily be sent to a laboratory, others cannot. In many cases FIELD CALIBRATION would be preferred.
Three major reasons for field calibration can be identified:
· Some flow meters are quite simple too large to calibrate in existing laboratories. Big size makes it inconvenient and sometimes impossible to remove the meters from their installation.
· Existing laboratories can offer calibration in only a limited number of liquids. If this liquid differ from the liquid used in process industries (such as paper pulp, chemicals, food products and high viscous oils) some meter types will be heavily influenced.
· Installation effects (how a flow meter is influenced by the flow profile of the liquid or other disturbances in the installation) do have an effect on many types of flow meters. On these types the calibration should therefore either be done on site, with the meter in its normal installation, or at ideal flow conditions.
1.2
Problem description and definition
The current situation is that there are problems to find suitable field calibration methods, for example for such installations of volume- and flow meters as can be found in process industry, petroleum refineries, district heating and water supply systems. The number of existing methods in practical use is very limited and furthermore each method has significant limitations, which sometimes are
undefined. The result is that many calibrations that really are needed either are not performed at all or performed in a manner that risk having unacceptable uncertainties. This is of course most
unsatisfactorily for such often very important measurements. Therefore, a guideline for finding the “right” method and knowledge of the limitations of available methods would be of great value for industry.
1.3 Realization
1.3.1 The
project
Nordtest has sponsored this project for a period of three years from 2002 to 2004. One reason for this relatively long time was that suitable field test possibilities had to be planned long time in advance. At early stage technical meetings was held with the companies Indmeas, Fortum (Birka Energi),
Panametrics, Preem and Södra Cell have been planned for the autumn 2002. Technical meetings and discussions have also been held with Justerdirektoratet in Norway, DOMS and FORCE in Denmark, NMi in the Netherlands and NMIJ in Japan.
1.3.2 Field
tests
Field tests for three different applications (oil, heat and paper pulp) were planned for this project. All have however met practical difficulties. These difficulties are mainly the same as the ordinary flow meter user, who is looking for a calibration method, is meeting, and should maybe therefore not be a surprise for the project members. At most sites there is simply no possibility to find a suitable “reference”. Other sites that we had in mind have been rebuilt, closed or are simply not longer available due to changes in production. Due to this fact, simulated laboratory tests have been performed instead.
1.3.3
Master thesis project
As a part of the research regarding field calibration of flow meters two master thesis project regarding digital communication with instruments were carried out. The reason for this is that many modern instruments, and especially those included in complex control systems, do not implement visual displays, pulse or analogue signals to transmit the instrument readings. Therefore procedures to access the instruments through different digital interfaces had to be examined, as well as possible limitations or error sources due to the digital communication.
1.3.4
Suppliers and manufacturers
To investigate the market a number of companies were asked in a questionnaire about their possibility to deliver instruments suitable for field calibration of flow meters. A workshop was held during 2003 to improve the communication with manufacturers and suppliers of flow meters and possible field calibration equipment.
1.4 Exclusions
There is a number of flow metering techniques that this project does NOT include. Equipment used for “service and repair” (with high uncertainty), “custody transfer” (traditional, verified methods) or any types of flow meters for air and gas are not mentioned.
2
Market survey
A market survey was performed during 2002/2003, searching for “Flow meters for field calibration purposes”. About 100 Nordic companies that supply different types of flow metering devices have been identified. A majority of these companies have been asked if they can supply some type of equipment for field calibration of flow meters. About 25% of the companies responded, and of these most can supply “ordinary” (or slightly modified) flow meters that can be used also for calibration. The suppliers were asked in a questionnaire about details such as media and material limitations, response time and similar. These answers have been matched with a number of data sheets into a table, presented in section 5 of this report.
No instruments primarily made for calibration of flow meters have been found, except traditional volume tanks and compact provers used for custody transfer. Several portable clamp-on ultrasonic meters are available. Of course these can be used for calibration but they are primarily manufactured as portable flow meters.
The discussions with different manufacturers of flow meters in general and clamp on in specific, have so far not indicated that any dedicated field calibration instruments exist. There is however openings for development depending on realized market needs for such equipment.
In addition, there are a number of companies that offer to do measurements with clamp-on ultrasonic meters or other commercially available equipment (such as dP-meters on control valves). These companies are mainly working with maintenance and have not been included in the survey. In the market survey three possible types of equipment, suitable for field calibration of flowmeters could be identified:
· Equipment for tracer measurements · Clamp-on ultrasonic flowmeters
· Master meters (“ordinary” flowmeters for portable use)
The number of available ”master meters” is very high, and suppliers of these are not listed here. The number of available clamp-on ultrasonic flowmeters is more limited, and identified companies are listed in the table below.
Represented by… Brand
in Sweden in Norway in Finland in Denmark Aaliant (USA) TemFlow Control
AB
Sigurd Soerum A/S Oy Saato AB Aaliant Badger (USA) PAAB Automation Sigurd Soerum A/S Bencon OY V. Loewener A/S Controlotron (USA) Sensortech Sweden
AB Flow Teknikk AS Sintrol OY Rikotech Eesiflo (USA)
Endress & Hauser (CH) Endress+Hauser
AB Endress+Hauser A/S Metso Endress+Hauser Oy
Endress+Hauser A/S
Flexim (D) Mobrey AB PAG-Automasjon
A/S Oy Autrol AB Fuji (Japan) Palmstierna AB
Katronic (UK)
Krohne (D) Gustaf Fagerberg AB
KROHNE Instrumentation AS
Oy Tecalemit AB G. Fagerberg A/S Micronics (UK) Satron AB Axflow Norsk
Pumpe AS Sarlin Process Automation Alfa Instruments Panametrics (USA) GE Panametrics
Sweden Pemac A/S Enprima Ltd. HH Instruments A/S
Polysonics (USA) Thermo Electron
Oy Pulsar (USA)
Ultraflux (F) Suomen Putkisto
Tarvike Oy
Note! Only companies responding to the questionnaire are listed. Please observe that some brand or supplier might be missing.
Parallel to this survey, suitable field methods and resources in existing “CMC tables” from European institutes have been studied. These tables list national resources for different calibration facilities worldwide. However, no field calibration resources have been included in these tables.
3
Workshop with instrument suppliers
In October 2003 a workshop took place in Boras. To this, major Nordic suppliers of flow instruments were invited. About 10 companies accepted the invitation and participated. The workshop focused on finding new ways for possible cooperation, calibration methods and customer requests in this area.
4
Existing Calibration methods
4.1 Tracer
methods
One of the known methods for field calibration of flow meters is the tracer method. This method is used in a lot of different applications, measuring everything from blood inside the human body to wastewater. In the process industry area, the largest private company in this field (in the Nordic countries) is Indmeas in Finland. Indmeas, who are accredited to ISO2975, have a lot of experience in this method and have also published some different investigations of the methods. They have about 15 people working with field calibration and claim to have performed more than 7000 measurements. They mainly work with tracers and transit time- and dilution methods. The accuracy is claimed to be in the region of 0,8 to 2%, depending on the application.
Another company working with similar methods is the Swedish Geosigma. They focus on
measurements of underground streams and water movements and their main customers are building and construction companies. However, they also can perform measurements of liquid flows in conduits. Examples of other companies that are using different types of tracer methods are Thermochem Laboratory and Consulting Services (USA) and Tru-Tec Services Inc (USA). There are also a number of companies that internally use different methods to perform flow measurements and calibration of flow meter equipment. Mainly these are companies that have a chemical lab available for test analysis. Examples are paper mills and waste water plants. One common method is to measure dilution of lithium. Tracer methods require special equipment and experienced personnel, and its use is relatively expensive.
4.1.1
The transit time method
In this method the travelling time of an injected portion of isotopes between two sensors is measured. This time is then recalculated to volume flow. In this calculation the distance and the pipe size (mean value of inner area) must be known between the sensors.
4.1.2
The dilution method
The tracer-dilution method is capable of measuring both open channel and closed conduit flow. Either isotopes, salts or dyes may be used as tracers. The tracer-dilution method consists of adding a known, strong concentration of tracer solution at a constant rate to the flow. Then, by chemical analysis, the downstream-diluted uniformly mixed concentration is measured. No measurements of flow section geometry or reach distance are required because the total flow is measured directly.
4.2
Clamp-on ultrasonic flow meters
The earliest paper on clamp-on ultrasonic flow meters was dated 1954, showing an instrument for plastic pipes capable to measure liquid velocities between 1 and 100 cm/s. The instrument claimed a linearity of 2%. 1964 a more practical and commercially available flow meter was introduced in Japan. Today, liquid flow is measurable by clamp-on equipment for industrial conduits as small as 10 mm and we can select among a lot of different models and features, divided in three basic working principles.
4.2.1
The transit time principle
To measure transit time, or “time of flight”, between two diagonal positioned sensors in a pipe is the most common method among available clamp-on ultrasonic flow meters.
The picture above shows the sensors A and B, where the difference between transit times tAB and tBA
gives the flow velocity. For correct measurements and recalculation from velocity to flow some fact has to be known. The area of the pipe is most important, but also media, pipe material, pipe thickness and eventual coating will influence the result.
Also in this method a transit time between sensors at a known distance is measured. Here, two ultrasonic detectors are sensing “in homogeneities” in the flow, causing a specific signal pattern. This pattern is then recognised at the downstream sensor and the flow can be calculated based on transit time, distance and pipe dimension.
4.2.3 The
doppler
method
A continuous signal at a fixed frequency is transmitted into a flowing fluid. Reflections from “objects” (particles, bubbles, vortexes) passing cause a frequency shift. A detector is receiving this “reflected” frequency and the difference can be recalculated to flow velocity.
4.3
Calibration using “built in” equipment
At some industries they regularly use the possibility to check their instruments against each other in a running process. For example two flow meters installed in series can be compared to each other. At oil refineries also more sophisticated equipment (such as ball- or piston provers) can be permanently installed. Another possibility is to check a flow meter against changing surface level in a storage tank. To use equipment like a level meter and a tank can be easy, but it is often rather complicated to calculate the uncertainty of the result.
4.4 Standardisation
in
the
field of meter calibration
4.4.1 Tracer
methods
ISO 2975-1:1974 Measurement of water flow in closed conduits -- Tracer methods -- Part 1: General
ISO 2975-2:1975 Measurement of water flow in closed conduits -- Tracer methods -- Part 2: Constant rate injection method using non-radioactive tracers ISO 2975-3:1976 Measurement of water flow in closed conduits -- Tracer methods --
Part 3: Constant rate injection method using radioactive tracers ISO 2975-6:1977 Measurement of water flow in closed conduits -- Tracer methods --
Part 6: Transit time method using non-radioactive tracers
ISO 2975-7:1977 Measurement of water flow in closed conduits -- Tracer methods -- Part 7: Transit time method using radioactive tracers
An application of ISO 2975 can be found in NORDTEST METHOD NT VVS 082, (liquid flow metering installations: radioactive tracer transit time method, in situ calibration). Ref: UDC 532.57 ISO 4053 “Measurement of gas flow conduits - Tracer methods” has been cancelled during 2003 (part 1 and 4).
4.4.2 Ultrasonic
methods
ISO/TR 12765:1998 Measurement of fluid flow in closed conduits -- Methods using transit-time ultrasonic flow meters
4.4.3 Related
topics
ISO/TR 3313:1998 Measurement of fluid flow in closed conduits -- Guidelines on the effects of flow pulsations on flow-measurement instruments ISO/TR 7066-1:1997 Assessment of uncertainty in calibration and use of flow
measurement devices -- Part 1: Linear calibration relationships ISO 7066-2:1988 Assessment of uncertainty in the calibration and use of flow
measurement devices -- Part 2: Non-linear calibration relationships ISO 8316:1987 Measurement of liquid flow in closed conduits -- Method by
collection of the liquid in a volumetric tank
ISO 11631:1998 Measurement of fluid flow -- Methods of specifying flow meter performance
5
Practical problems
5.1
Access to meters
Some type of access is required for all types of calibration systems. Master meters has to be connected is series, which require flanges, valves and flexible connections. Details about this are described in section 6.1 in this report. But also clamp-on meters require some type of access. Eventual pipe
insulation has to be removed and an undisturbed straight pipe section is required for best performance. Measurement with tracers require some type of open connection where the tracer liquid can be
inserted, and downstream of this point (after mixing) an straight undisturbed section for mounting of sensors or another open connection for sample outlet.
5.2 Signals
To be able to calibrate a flowmeter a signal must be available. This signal is then compared with the signal from the selected calibration equipment. Typical output signals from flowmeters are analogue (0/4-20 mA), pulse (frequency) or visual (display/counter).
When looking at these interfaces from a calibration point-of-view there are mainly three important features.
- Response time
At what frequency are the signals transmitted and what is the response time at fast changes in the measured variable. This is very important when synchronizing the test object with the test equipment.
- Resolution
The resolution of the digital representation of the measurement value will affect the uncertainty of the measurement.
- Security
How is the transfer guaranteed? Is the transferred data submitted to any kind of internal control such as checksums etc? How is the robustness against electrical noise?
Many modern instruments, and especially those included in complex control systems do not
implement traditional instrument readings, but instead computer technology and different “Field Bus” systems are used. Therefore procedures to access the instruments through different digital interfaces two master thesis projects were initiated within this project, to examine possible limitations or error sources due to the digital communication.
5.2.1 HART
The first master thesis dealt with calibration of flow meters using the HART protocol. HART is a serial master-slave communication protocol that transfers data by superimposing an AC- signal on the analogue DC- current signal. When used between only two units (point-to-point) the analogue signal can be used simultaneously with the digital transfer, but when used in a network (up to 15 units) only digital communication is possible. The master thesis resulted in a “HART spy”, i.e. a device that can be used to connect a standard PC to a HART bus. Software was also written to transform the acquired data to the correct format and to integrate the data with the available measurement software normally used at SP. This device makes it easy to access the data from an instrument communicating with the HART protocol. During the project it was found that communication through HART is relatively slow, about 1-2 updates per second, which can contribute to synchronization problems when measuring fluctuating flows. It was found that increasing the baud-rate of the host PC made no improvements in the response time. An exaggerated baud-rate even caused communication failures.
5.2.2 Profibus
The second master thesis examined response times for the Profibus fieldbus. This protocol is a serial master- slave communication protocol with up to 126 units per net and exists in four different variants depending on its usage. The variant examined in the master thesis was the Profibus DP, which is mainly used for workshop automation and the transfer is made through RS485. If the bus where the calibration is to be made supports the Profibus DP-V2 protocol and has a clock-master for
synchronization each measurement can be given a time-stamp with an accuracy of < 1ms, which makes it very suitable for calibration purposes. If no time-stamp possibility exists, the response time depends on the number of units connected to the net. The master thesis project performed
measurements of the response time for a small net with one master and two slaves, and the response times were well within acceptable limits (always less than < 40 ms at all baud-rates). Tests were also performed when the net was submitted to electrical noise. The communication was able to handle the disturbance well, although the response time increased slightly, especially for the higher baud-rates. These tests were performed at a very limited net, and can therefore not be used to guarantee certain response times, but the form a very strong indication that the Profibus is more than fast enough to perform field calibrations.
6
Evaluation of methods
Some different laboratory tests have been performed to evaluate different methods for calibration, using master meters, clamp-on ultrasonic meters and tracer methods. Of course, these tests do not fully answer questions regarding factors like sensitivity to media. The results can be divided in two parts. One part indicates best possible results under optimum conditions. In the other part attempts have been made to simulate “normal field difficulties”, such as unknown pipe dimensions, short straight pipe sections and varying temperature.
6.1
Test 1 – Magnetic inductive flow meter as master
One inductive flow meter (Krohne X1000 /IFC110 F) was supplied with fixed stainless steel pipes at the inlet and outlet section. The inlet section was equipped with a flow straightener consisting of a perforated plate and an anti swirl device (two plates mounted as a cross). At the end of the pipes there was connectors for flexible hoses.
The complete package with meter, straightener and pipes was then calibrated with the hose is three different positions. Flow rate l/min Temp. °C Pressure kPa
Pos. Error of indication % Range % Uncertainty % 602 20 218 1 +0,17 0,01 ±0,10 594 22 197 2 +0,48 0,02 ±0,10 609 23 230 3 +0,99 0,01 ±0,10
Installation /position of the inlet section
Conclusions: Even if an inductive flowmeter is equipped with flow straightener and fixed inlet- and outlet sections it is difficult to reach an uncertainty less than 0,5%.
Pos. 1
Long straight run
Pos. 2 One 90º bend
Pos. 3
6.2
Test 2 – Ultrasonic clamp-on flowmeter at different pipe
materials
An ultrasonic clamp-on flow meter from Controlotron of type UNIFLOW was used on four different pipes:
- one made of stainless steel with well known dimensions
- two made of carbon steel (in two sizes), one with inside corrosion (▲ in graph). - one made of plastic (PEN)
The result is presented in the graph below. Exact information of pipe sizes and wall thickness, as well as eventual corrosion on the inside was not known. Both water and petroleum products were used. All tests were made at ambient temperature.
Error of indication (%) = x 100 value Reference value Reference -flow) c (volumetri reading Meter Pipe material -8 -6 -4 -2 0 2 4 6 8 10 12 0 2 4 6 8 10 Velocity (m/s) Error of indication (%) Stainless steel P E N Steel Steel
Conclusions: Information and knowledge of pipe material, corrosion, coating and dimension is important and necessary when using a clamp-on flowmeter. With plastic pipe or corroded steel pipes it is difficult to reach an uncertainty less than 5%.
6.3
Test 3 – Ultrasonic clamp-on flowmeter at different liquid
temperature
An ultrasonic clamp-on flow meter from Panametrics of type PT868 was used at two different temperatures. Only one, well known, pipe (diameter 50 mm, stainless steel) was used. The result is presented in the graph below. The liquid was water.
Error of indication (%) = x 100 value Reference value Reference -flow) c (volumetri reading Meter Temperature -8 -7 -6 -5 -4 -3 -2 -1 0 0 0,5 1 1,5 2 2,5 Velocity (m/s) E rror of indication (%) 80 ºC 15 ºC
Conclusions: Temperature seems to influence the result quite much. According to datasheet and other tests the shift in this test is a bit too large. Discussions with manufacturers give no answer why. One reason, beside from the actual temperature influence, might be that hot water contains more bubbles.
6.4
Test 4 – Ultrasonic clamp-on flowmeter at different pipe
sizes
An ultrasonic clamp-on flow meter from Panametrics of type PT868 was used at four different pipe sizes. The result, showing pipe inner diameter and error of indication is presented in the graph below. The liquid was water at ambient temperature and the flow velocity 2 m/s. The pipes has not been measured, instead data given by the pipe manufacturer has been used.
Error of indication (%) = x 100 value Reference value Reference -flow) c (volumetri reading Meter Pipe size -10 -8 -6 -4 -2 0 2 4 0 20 40 60 80 Inner diameter (mm) Error of indication (%) 2 m/s
Conclusions: Manufacturers in general seems to expect problem at diameters less than around 20 mm. In this test we have unexplained errors also for large diameter. Length on inlet and outlet sections has been sufficient, according to the installations manual, in all cases. However, the inlet sections were relatively longer for the smaller pipes. Installation recommendations (regarding length of straight undisturbed pipe) from the manufacturer sometimes are insufficient.
6.5
Test 5 – Ultrasonic clamp-on flowmeter at different
positions
The following measurements were all made on water at a temperature of 21°C with the meter mounted on a pipe with an outer diameter of 54.0 mm and a wall thickness of 2.0 mm. Each measurement is compounded of the mean value of 36 instrument readings. The flow velocity was 0,3 m/s at all positions. The ultrasonic clamp-on flow meter was a Panametrics of type PT868.
Position Distance after pipe bend in multiples of pipe diameter Error of indication % Range (max-min reading) % 1 ∞ 2,1 ±0,8 2 5 -6,6 ±1,5 3 10 -2,3 ±0,2 4 10 -2,0 ±1,9 5 5 -2,2 ±1,3 6 5 -8,5 ±1,5 7 10 0,7 ±0,2 8 10 -2,7 ±0,2 9 10 -0,7 ±0,3
Conclusions: Just as expected, mounting close to a bend results in higher error and more instability. Different angles at the same position also affect the result. Compensation for mounting close to a disturbance seems not possible since the effect is varying.
6.6
Test 6 – Using a ultrasonic clamp-on flowmeter as master
for calibration
6.6.1 Measurement
methods and procedures
Two reference meters, one with pulse output signal (meter 1) and one with both pulse output (meter 2) and mA-output signal (meter 3) has been calibrated using a clamp-on ultrasonic flow meter. This calibration was repeated 3 times (at different positions and with different settings of the ultra sonic flow meter). The reference meters were in the same position during all tests. The measurements were performed by personnel from G E Panametrics. During all tests the analogue output from the
ultrasonic meter was used. The signal was compared with the output from the reference meters in an external system (part of rig VM3). Two GE Panametrics type PT868 were used.
6.6.2 Test
rig
set-up
Meter 1: Endress+Hauser Autozero DN100, pulse output. (ID: 600023)
Meter 2: Krohne X1000 DN80/ SC100 AS, pulse output. (ID: 600325 + 600328) Meter 3: Krohne X1000 DN80/ SC100 AS, 4-20mA output. (ID: 600325 + 600328)
Test section Ball prover (3500 l) Meter 2-3 Meter 1 Storage tank (60 m3) Drain A B C C 1 m 6 m
6.6.3
Error of indication
Flow rate l/min Pres. MPa Temp. ºC Meter 1 % Meter 2 % Meter 3 % Position A, fact. cal. * 1190 0,1 19,1 +3,7 +3,4 +3,4Reference -0,23 -0,54 -0,52
Position A, fact. cal. 620 0,1 19,2 +4,4 +3,7 +3,8
Reference -0,07 -0,72 -0,67
Position A, fact. cal. 300 0,1 19,6 +0,7 -0,3 -0,1
Reference +0,04 -0,73 -0,55
Position A, local cal. ** 1190 0,1 19,1 -0,2 -0,5 -0,5
Reference -0,21 -0,57 -0,58
Position A, local cal. 600 0,1 19,0 +0,5 -0,1 ±0,0
Reference -0,07 -0,68 -0,62
Position A, local cal. 300 0,1 19,3 -2,0 -2,8 -2,6
Reference +0,04 -0,73 -0,55
Position B, fact. cal. 1200 0,1 19,1 +3,5 +3,2 +3,2
Reference -0,23 -0,52 -0,52
Position B, fact. cal. 620 0,1 19,1 +4,2 +3,6 +3,6
Reference -0,07 -0,70 -0,64 *) factory calibration = electronics and sensors not delivered in pair
**) local calibration = wet calibration with electronics and sensors in pair. This calibration was made at an accredited laboratory in Sweden.
6.6.4 Pipe
measures
Since an ultrasonic clamp-on flow meter primarily measures the speed of the flow, it is essential for the instrument to “know” the pipe measure; diameter and thickness (internal area). In all results presented on previous pages the instruments has been programmed with the pipe measures that were printed on the pipes by the manufacturer. “Nom. area” in the table below is also based on this marking.
After the calibration the pipes were checked and a reference measurement was performed inside the pipe. This indicates the inner diameter close to the ends of the pipe section. At these points the diameter was measured on 4 positions and the reference area was calculated. The uncertainty in the diameter measurement is < 0,05 mm but the roundness of the pipe is not included.
Pipe Marking Nom. area
mm2 Reference area mm2 Dif. %
Pos A DMV F-ASTM A312 114,3 x 6,02 8213,0 8197,3 0,2
Pos B HAATO AISI316 104 x 2 7854 7829 0,3
6.6.5
Results – comparison including stated uncertainties
Different measurement and calibration equipment have different advantages and disadvantages. In this report we only look at measurement errors and uncertainties. One way of comparing results from instruments with different stated uncertainties is to use the “En-method”. This method is commonly
use by the European Accreditation organisation (EA). En > 1 indicate that the difference to the
reference is too large OR that the stated uncertainty is to small.
) U (U E -E E 2 2 ref ref n + =
Meter 1 Meter 2 Meter 3
Flow rate l/min Error % Uncert. % En Error % Uncert. % En Error % Uncert. % En A1 1190 +3,7 5 0,8 +3,4 5 0,8 +3,4 5 0,8 Ref. -0,23 0,2 -0,54 0,2 -0,52 0,2 A1. 620 +4,4 5 0,9 +3,7 5 0,9 +3,8 5 0,9 Ref. -0,07 0,2 -0,72 0,2 -0,67 0,2 A1 300 +0,7 5 0,1 -0,3 5 0,1 -0,1 5 0,1 Ref. +0,04 0,2 -0,73 0,2 -0,55 0,2 A2 1190 -0,2 5 0,0 -0,5 5 0,0 -0,5 5 0,0 Ref. -0,21 0,2 -0,57 0,2 -0,58 0,2 A2 600 +0,5 5 0,1 -0,1 5 0,1 ±0,0 5 0,1 Ref. -0,07 0,2 -0,68 0,2 -0,62 0,2 A2 300 -2,0 5 -0,4 -2,8 5 -0,4 -2,6 5 -0,4 Ref. +0,04 0,2 -0,73 0,2 -0,55 0,2 B 1200 +3,5 5 0,7 +3,2 5 0,7 +3,2 5 0,7 Ref. -0,23 0,2 -0,52 0,2 -0,52 0,2 B 620 +4,2 5 0,8 +3,6 5 0,8 +3,6 5 0,8 Ref. -0,07 0,2 -0,70 0,2 -0,64 0,2 E= error of indication (%) U= uncertainty (%)
6.7
Test 7 – Using the tracer method for calibration
6.7.1 Measurement
methods and procedures
Two reference meters, one with pulse output signal (meter 1) and one with both pulse output (meter 2) and mA-output signal (meter 3) has been calibrated using tracer injection methods (both transit time and dilution). This calibration was performed with the tracer sensors mounted in two different pipe sections (6 or 1 meter from a 90º bend) and with three different settings of the calibration equipment. The reference meters were in the same position during all tests. Personnel from Oy Indmeas
performed the measurements and the results have been presented in calibration certificates.
6.7.2 Test
rig
set-up
See section 6.6.2
6.7.3
Error of indication – transit time method
Flow rate l/min Velocity m/s Pres. MPa Temp. ºC Meter 1 % Meter 2 % Meter 3 % Position A 1210 2,6 0,1 22,0 -0,3 -0,8 -0,8 Reference -0,23 -0,72 -0,72 Position A 600 1,3 0,1 22,3 -0,3 -1,0 -0,8 Reference -0,09 -0,86 -0,80 Position A 300 0,6 0,1 21,0 -0,3 -1,2 -0,6 Reference +0,02 -0,94 -0,78 Position A * 300 0,6 0,1 20,5 -0,1 -1,1 -0,5 Reference +0,02 -1,04 -0,83 Position A ** 300 0,6 0,1 19,7 -0,1 -1,1 -0,5 Reference +0,02 -1,02 -0,86 Position B 1200 2,6 0,1 20,3 +0,6 +0,2 +0,3 Reference -0,23 -0,62 -0,62 Position B 300 0,6 0,1 20,0 -0,1 -1,1 -0,5 Reference +0,02 -0,99 -0,82 *) special tracer mixture with Na2SO4 and H2SO4
6.7.4
Error of indication – dilution method
Flow rate l/min Velocity m/s Pres. MPa Temp. ºC Meter 1 % Meter 2 % Meter 3 % Position C 1200 2,5 0,1 23 -0,3 -0,6 -0,6 Reference -0,21 -0,48 -0,52 Position C 750 1,6 0,1 22 -0,6 -1,0 -1,1 Reference -0,07 -0,56 -0,56 Position C 350 0,7 0,1 19 -0,7 -1,6 -1,2 Reference +0,04 -0,73 -0,616.7.5 Pipe
measures
Since tracer injection methods primarily measure the speed of the flow, it is essential for the operator to “know” the pipe measure; diameter and thickness (internal area). In all results presented on
previous pages the operator has used pipe measures that were measured (from outside of pipe) on site by the operator himself. After the calibration the pipes were checked and a reference measurement was performed inside the pipe. This indicates the inner diameter close to the ends of the pipe section. At these points the diameter was measured on 4 positions and the reference area was calculated. The uncertainty in the diameter measurement is < 0,05 mm but the roundness of the pipe is not included.
Pipe Measured area
mm2 Reference area mm2 Diff. %
Pos A DMV F-ASTM A312 8185,7 8197,3 -0,1
Pos B HAATO AISI316 7828,9 7829,3 ±0,0
6.7.6
Results – comparison including stated uncertainties
Different measurement and calibration equipment have different advantages and disadvantages. In this report we only look at measurement errors and uncertainties. One way of comparing results from instruments with different stated uncertainties is to use the “En-method”. This method is commonly
use by the European Accreditation organisation (EA). En > 1 indicate that the difference to the
reference is too large OR that the stated uncertainty is to small.
)
U
(U
E
-E
E
2 2 ref ref n+
=
6.7.6.1
Transit time method
Meter 1 Meter 2 Meter 3
Flow rate l/min Error % Uncert. % En Error % Uncert. % En Error % Uncert. % En A 1210 -0,3 0,8 -0,1 -0,8 0,8 -0,1 -0,8 0,8 -0,1 Ref. -0,23 0,2 -0,72 0,2 -0,72 0,2 A 600 -0,3 0,8 -0,3 -1,0 0,8 -0,2 -0,8 0,8 ±0,0 Ref. -0,09 0,2 -0,86 0,2 -0,80 0,2 A 300 -0,3 0,8 -0,4 -1,2 0,8 -0,3 -0,6 0,8 +0,2 Ref. +0,02 0,2 -0,94 0,2 -0,78 0,2 A* 300 -0,1 0,8 -0,1 -1,1 0,8 -0,1 -0,5 0,8 +0,4 Ref. +0,02 0,2 -1,04 0,2 -0,83 0,2 A** 300 -0,1 0,8 -0,1 -1,1 0,8 -0,1 -0,5 0,8 +0,4 Ref. +0,02 0,2 -1,02 0,2 -0,86 0,2 B 1200 +0,6 1,3 0,6 +0,2 1,3 0,6 +0,3 1,3 +0,7 Ref. -0,23 0,2 -0,62 0,2 -0,62 0,2 B 300 -0,1 1,2 -0,1 -1,1 1,2 -0,1 -0,5 1,2 +0,3 Ref. +0,02 0,2 -0,99 0,2 -0,82 0,2 E= error of indication (%) U= uncertainty (%)
6.7.6.2 Dilution
method
Meter 1 Meter 2 Meter 3
Flow rate l/min Error
%
Uncert.
% En Error % Uncert.% En Error % Uncert. % En
C 1200 -0,3 0,9 -0,1 -0,6 1,0 -0,1 -0,6 0,9 -0,1 Ref. -0,21 0,2 -0,48 0,2 -0,52 0,2 C 750 -0,6 0,7 -0,7 -1,0 0,8 -0,5 -1,1 0,7 -0,7 Ref. -0,07 0,2 -0,56 0,2 -0,56 0,2 C 350 -0,7 0,8 -0,9 -1,6 0,8 -1,1 -1,2 0,8 -0,7 Ref. +0,04 0,2 -0,73 0,2 -0,61 0,2
6.8
Evaluation at other institutes
6.8.1
Ultrasonic clamp-on meters
Performances of clamp-on meters are in general claimed to be between around ±5%. Some manufacturers claim down to ±1%. Independent tests shows results somewhere in between. One problem with the stated uncertainty is that it is often valid only for a “perfect” installation, which is hard to find. A number of evaluation and research projects regarding ultrasonic flowmeters (both in-line and clamp-on) have been made recently
6.8.1.1
Cranfield University, UK
One comprehensive study has recently been performed at Cranfield University. In this report effects from pipework, mounting and user training has been investigated. Pipe material (and inner surface roughness) also seems important; pipe wall roughness can cause an error of up to 4% in very rough pipes. Because the overall performance achievable in the field depends on the care with which the operators prepare the pipework and mount the transducers coupled with an inability to provide
traceable calibrations of clamp-on meters in the field it is difficult to obtain figures for the uncertainty, repeatability and reproducibility of such flowmeters in the field. Other important sources of
disturbances are effects from air bubbles and solids.
6.8.1.2 NIST,
USA
Within NIST's “Ultrasonic Technology Assessment Program to Improve Flow Measurements” one report presents results of the first phase of a three-phase effort conducted at NIST to assess ultrasonic technology for improving flow measurement. Each of the three phases is planned to contain results from: (a) testing commercially available, travel-time, clamp-on type ultrasonic flow meters, (b) modelling the pipe flows involved, and (c) producing computer simulations for wide variations of the arrangements of these meters.
The first phase test results show that most of meters tested had errors in the range from 0% to 3%, relative to NIST’s static gravimetric flow standards. The worst-case error was –14%. Results also showed that these meter manufacturers have progressed well in correcting historical problem areas associated with “removereplace” variations and with “zero-flow” set-up requirements in order to attain specified performance.
The flow modelling results indicate that for the conditions tested, very long lengths of constant diameter piping are required to produce ideal, fully developed flow conditions. The computer simulations of ultrasonic metering techniques indicate the consequences of the software strategies used to process final results and they indicate the trends in performance as flow conditions vary. These simulations indicate that, while the manufacturers of the tested meters are compensating for pipe flow distributions, the compensations are not sufficient, and the trends shown with flow rate are frequently opposite to that shown by the simulations.
6.8.1.3 Karlsruhe
University,
Germany
This is one of few projects where a cross correlation flow meter has been investigated.
The meter exploits only the natural inhomogeneities of a one phase flow without introducing any bluff bodies. Measurements with clamped on and direct contact transducers was carried out in the Reynolds number range 4000–250 000 and for various pipe diameters and materials. The measurement error was found to be less than 3%.
Performance test result on a cross correlation meter. Test performed at Karlsruhe University
7
Preliminary guide for the selection of methods
There is a long list of questions to ask when selecting a suitable method for field calibration of flowmeters. The very first question is if the calibration is best done in the field or in a laboratory. In general, a flowmeter with good performance and low sensitivity to flow profile and other installation effects should be calibrated in a lab.
Next question is probably what the required performance (accuracy) is. With this info it is possible to search for a matching calibration method. As mentioned, the number of existing methods is however quite limited. In this selection guide we have included five methods. There are also practical questions regarding connection possibilities, available space, environment (Ex-zone?), health risk
(poisonous/hot liquids), material (corrosion/contamination) and available output signals. The following methods are included in the selection chart:
Volume standard, or portable weighing tank (V) Master meter, “high quality” (M)
Clamp-on ultrasonic meter (C) Transit time tracer method (T)
Other methods incl.“low quality” master meter (X)
Approx. dimension Media Temperature
Required
performance <40 <100 <300 >300 Clean solids With bubbles With High visc. Amb. High Low ± 10% X V M C T X V M C T X M C T X C T X V M C T X M (C) T X M (C) T X M C T X V M C T X M C T ± 5% X V M C T X V M C T X C T X C T X V M C T X M T X T X M (C) T X V M C T X M C T ± 1% M V T V M T T T V M T (T) (T) M (T) V M T M T ± 0,5% M V M V V M M V M ± 0,2% (M) V V V (M) V
Combination of methods can in some cases improve the performance. For example, some types of master meters can have a reduced uncertainty if calibrated “on-site”. A “high quality”-master meter is not only a good instrument, but it has also a “history” with calibrations, documented effect from fluid properties (like density, viscosity and temperature) and a good, proven repeatability.
7.1 Volume
standards
Tanks used as volume standards should have a resolution in the volume reading, suitable for the required uncertainty. They should of course also have a traceable calibration. In field use most often the “standing start and stop”-method is used, and the total volume of the tank must be large enough to prevent influence from the start- and stop sequence. It is also important to arrange the inlet to the tank in such a way that the “interface” between two fillings is accurately defined. A sight glass is useful to detect air and gas bubbles. For low uncertainties it is necessary to make temperature compensations for both tank temperature (deviation from calibration) and liquid temperature (deviation between tank and meter).
In some cases devices like piston provers and ball provers can replace volume standards. The advantage is a higher possible flowrate and a more time efficient calibration procedure.
7.2 Master
meters
Most types of flow meters can only be used in a specified type of liquid at nominal flow range and within a given range of temperature and pressure. These conditions can most often also be achieved if the meter is connected with flexible hoses to a process line. For best performance it is recommended to combine the meter with fixed inlet and outlet sections. In this way most types of “standard” flow meters can be considered as portable.
A “master meter” is a portable meter that can be used in series with another (permanently installed) meter for calibration purposes. Some aspects are very important when selecting and using a flow meter as a “master”;
1. Selection of type and size is even more important than for a normal installation, since a very good performance is required from a master meter.
2. Good repeatability is valuable. Even if the linearity is not so good, it is possible to make corrections and compensations to improve the results. But only if the repeatability is good enough!
3. Sensibility to temperature, pressure, viscosity, density and other influencing factors must be kept as small as possible. Or known and compensated for.
4. Sensibility to flow profile and installation effects can be a big problem. Therefore most types of meters require fixed inlet and outlet sections (straight pipes) when used as master meters. Sometimes a flow strainer is required to reduce the effect from the connection and the position of the flexible hose. A control valve downstream of the meter can be used to adjust pressure and flow rate (this valve is absolutely necessary when the master is placed
downstream the test object, with open outlet to a drain). Since rubber hoses connect the meter, an electrical earth connection cable can be needed.
5. The connection to the process line (and the flow meter to be tested) has to be done carefully. If a valve is used to redirect the flow, this valve must be leak proof. A so called “block and bleed” valve (or three valves mounted according to the figure) makes it possible to check for leaks. The distance between the two meters must not be to large and it must be possible to evacuate all entrapped air and gas from the system. In heated systems also the master meter loop normally has to be heated.
6. When the master has been installed and all flow is redirected it is time to compare the output signals from the two meters. If mechanical counters or electronic displays are used these must be synchronised. Resetable totalizers can be activated simultaneously. Flow indicators may be difficult to compare “manually”, especially if the flow rate is fluctuating. In such cases synchronised cameras can be used, to “freeze” the momentary indication of both displays. Electronic flow meters, with pulse output signals, can be compared by using a double counter with synchronised start, stop and reset. If a timer is synchronised together with the counters also mean flow rate can be calculated. In both cases the resolution of the output (number of decimals / pulse value) is important and the time for each test must be selected so that a sufficiently large volume has passed the meters.
Example: A display with 0,1 litre in resolution and a test volume of 100 litre results in an uncertainty of 0,1% due to the reading of the display.
p t Undisturbed outlet Undisturbed inlet Control valve Flow strainer Meter Master Meter on test ”Block and bleed”
valve arrangement Counter 1 0000 Counter 2 0000 Timer 0000 Clock oscillator Pulse signal Pulse signal Start /
7.3
Clamp-on ultrasonic meter
Practical aspects to consider when using clamp-on meters are media, temperature, available
undisturbed pipe sections, pipe material (corrosion) and possibilities to remove eventual insulation or coating. Liquids containing solid particles can’t be measured, nor liquids with air or gas bubbles. Knowledge of the instrument is important! There are many settings that have to be correctly done before measurements can be performed.
7.4
Transit time tracer method
Advanced and special equipment is needed, why probably this method require assistance from other companies. Guidance is available in ISO 2975 and NORDTEST method VVS082. A calibration is rather time consuming and need some preparations. To achieve lowest possible uncertainty there are some practical aspects like availability of suitable injection places (with downstream mixing) and a measuring section with a well defined pipe area.
7.5 Other
methods
Other methods can be a comparison to an existing storage tank, comparison to other flowmeters connected in series, using tanks on road trucks or ships and similar. In all cases analyse of the measurement uncertainty has to be performed individually, and in these cases it is even more important with an uncertainty calculation covering all aspects.
8
In the future…
8.1
Ongoing research at institutes
Contacts have been taken with Force and DOMS in Denmark, Justerdirektoratet in Norway, NEL in UK, NMi in the Netherlands, KRISS in Korea and NMIJ in Japan. DOMS has mainly resources and interests in ”traditional” custody transfer calibration (standard volume tanks), but likes to “follow the results of the project”. The same situation applies at Justerdirektoratet but they see an increasing request for this type of calibration and would like to continue the discussions and to arrange a meeting. NEL, KRISS and NMIJ have also interest in this area, but for the moment they focus on research for new “measuring methods”. Some universities have research in areas with close
connections to this project. Examples are University of Tampere (fluid dynamics in liquids with high concentration of solids), Cranfield University (general flow measurement subjects) and Luleå Tekniska Universitet in Sweden (ultrasonic technique).
Only one research project that is specifically aimed for field calibration of flow meters has been found in the Nordic countries. This project, called “Flow Act”, was initiated at Chalmers University in Gothenburg in 1994. The principle is similar to the traditional transit time injection method, but in this cased based on pulsed neutron activation (from the outside of the pipe). Particles in the water are activated and decay with a half-life of 7.13 s. At the decay, gamma radiation is emitted, and downstream detectors outside the pipe detect this radiation. The project has met difficulties in the signal analysis, but is still ongoing. The accuracy of the evaluation methods used so far has been estimated to 5-10%.
Of course, traditional field methods for custody transfer are available, and some research projects are made on these methods to improve accuracy or reduce working time. Such works are not included in this project, as they are not aiming for industrial use.
International research on the subject has been found in England, Netherlands, USA, Japan and Mexico. Almost all these research projects have been focused on ultrasonic technique and clamp-on methods. The general opinion seems to be that this technique is the most promising.
Only a few research projects concerning injection transit time technique have been found.
Many researchers are working with calculations and computer simulations of fluid dynamics and the effect on flowmeters. One example is a project at NEL called "Ultrasonic Meters in Non-standard Flows". Typical pipe configurations can produce flow distortions that cause large meter errors. Compensation for distorted flow profile is a very complex problem. The result of this NEL-project is so far that sophisticated CFD calculations a prediction can be made within approximately +/- 2%. Therefore CFD can be used to evaluate installation effects when the magnitude of error is large relative to the ideal performance.
In some aspects also equipment using a laser doppler velocimeter (LDV) can be considered for field calibration. However, this is primarily a instrument for research and development and today it is rather costly and time consuming to use in field application.
8.2
Research at manufacturers
Research is of course also going on at different manufacturers. More and more in-line multi path ultrasonic flow meters are available. The American Petroleum Institute, API, has performed a test of seven commercially available multi path meters. We have asked three of the participating companies (Krohne, Caldon and GE Panametrics) if they see any possibility (and market) for a “high accuracy multi path clamp-on ultrasonic flow meter”. They have all answered “no” (for the time being). The American company Caldon has taken another approach to the problem, with the lack of field calibration methods. They are specialized in the nuclear industry, and therefore they also supply flow meters to nuclear plants. The need for high precision flow meters (and for calibration) is growing, due to higher energy prices and requirements from IAEA. Without suitable field calibration methods Caldon instead offer to build a copy of the actual pipe work in which the flow meter is normally installed. The calibration can then be made in this copy, and the idea is that the measured uncertainty then can be “transferred” to the real site. Such calibrations can be made at ARL in Boston. The costs are however extremely high.
8.3
Combination of methods and principles
To improve accuracy and/or possible applications two or more existing methods can sometimes be combined. For example, measuring both differential pressure and flow velocity results in information of both volumetric flow and mass flow. Combination of a volume standard and a master meter reduces the uncertainty drastically. Similar possibilities exist and this is an area where more research seems to be of value.
8.4 User
Training
The simple appearance of a flowmeter and the simplified principle often presented belies the fact that most meters are very sophisticated devices and that there are many complex interrelated phenomena involved in producing a measurement. Users who undertake the measurement must be aware of these and take them into account when undertaking the measurement. This is the case also for convenient and “easy to use” portable clamp-on ultrasonic flowmeters. It must be stressed that clamp-on meters should be installed by trained personnel. Newer models ARE better and safer (with internal checking facilities), but also these instruments require some type of training.
One important future step is therefore to encourage manufacturers, institutes, universities, schools and other organisations to offer more education and practical training in the field of flow measurement.
9
Uncertainty and traceability
The uncertainty for all internally made reference measurements in this report is less than 0,20 % (results presented in section 6.1 to 6.7). This uncertainty includes contribution from both reference meters (range of dispersion), method and equipment, and is stated as the standard uncertainty of measurement multiplied by the coverage factor k = 2, which for a normal distribution corresponds to a coverage probability of approximately 95 %. The standard uncertainty has been determined in
accordance with EA Publication EA-4/02 (formerly EAL-R2).
The total uncertainty for measuring results including comparisons with calibration methods is separately stated in result tables 6.6.5 and 6.7.6.
10 Conclusions
The scope for this project has been to evaluate available methods and produce a preliminary guideline for the selection of suitable calibration methods for flow and volume measurement on site, including limitations and guide for the uncertainties of the methods. The objectives have been met and the results are presented in the report.
An important result of the project work is that limitations of the existing, practically available, calibration methods have been confirmed.
There is a need for the development of improved methods for calibration of flow meters on site, without the need of breaking in to the pipe. Discussions have been initiated for the collaboration for such development projects.
A lot of evaluation and research project are for the moment working with in-line multi beam ultrasonic flowmeters. The availability and performance of these instruments have improved drastically the latest three years. It seems reasonable that these improvements will benefit also the clamp-on versions in a near future.
11 References
“On-site calibration of liquid flow meters using the radiotracer transit time method”, Kuoppamaki R.; Turtiainen H.
“Guidelines for the use of Ultrasonic Non-Invasive Metering Techniques”, NMS Flow Programme: Project KT12 (Cranfield University).
A clamp-on ultrasonic cross correlation flow meter for one-phase flow, Universit ¨ at (TH), Institut f ¨ur Meß- und Regelungstechnik, Karlsruhe, Germany.
“NIST's Ultrasonic Technology Assessment Program to Improve Flow Measurements”,
G. E. Mattingly, T. T. Yeh, Fluid Flow Group, Process Measurements Division, Chemical Science and Technology Laboratory, National Institute of Standards and Technology.
“Development of a scintillator detector…”, Fabio E da Costa, Margarida M. Hamda, Instituto de Pesquisas Energeticas, Sao Paulo, Brazil.
“Improving the accuracy of tracer flow-measurement techniques by using an inverse-problem approach”, L Le Brusquet and J Oksman, Measurement Department, Ecole Superieure d’Electricite, Gif sur Yvette, France.
“Clamp-on ultrasonic flow meters”, Sensors Aug 2001 - Larry Lynnworth, Panametrics, Inc. “Flow Measurement Guidance Note No. 22”, NEL.
“Study of using pulsed neutron activation for accurate measurements of water flow in pipes”, Per Linden, Chalmers Tekniska Högskola.
SP Measurement Technology SP REPORT 2004:24
ISBN 91-7848-998-9 ISSN 0284-5172
technical investigation, measurement, testing and certfi cation, we perform
research and development in close liaison with universities, institutes of technology and international partners.
SP is a EU-notifi ed body and accredited test laboratory. Our headquarters are in Borås, in the west part of Sweden.
SP Swedish National Testing and Research Institute
Box 857
SE-501 15 BORÅS, SWEDEN
Telephone: + 46 33 16 50 00, Telefax: +46 33 13 55 02 E-mail: info@sp.se, Internet: www.sp.se
