Winter 2009

A Quarterly Newsletter of the Instrumentation Testing Association

ITA Enews

Spring 2011

Instrumentation Testing Association  (ITA)


Toll Free Inside the U.S. and CANADA 1-877-236-1256

Global Phone: 702-568-1445


Email ITA


From the previous sections it is clear meters have to be verified in-situ and on a regular basis.   Pipe characteristics (sedimentation and calcification) do change with time and it cannot be assumed that swirl does not form as a pipe ages.  Work in the UK at the same site for several years has identified bias effects caused by sedimentation to be assessed.   Full verification includes site surveys, pipework assessments, profile measurements, and telemetry output data analysis as standard plus information specific to each site.  The benefits have clearly been demonstrated at many sites in the UK.   These benefits include:

· Leakage reduction: Periodic full verification quantifies zonal and district input more accurately

· Works balances are fully quantified: existing methods neglect some fluid dynamics effects

· Reduced chemicals usage: better uncertainty analysis allows optimum volume addition

· Better capital employment: better assessment correctly prioritises metering problems

· Future design upgrades: Experience can be used to design out known deficiencies

· Abstraction metering: More accurate measurement of volumes in accordance with licenses

· Economic maintenance: Full verification looks at piping and valving as well as metering

· Asset Management: RPS methods allow better management of assets by complete documentation

· Verification tools: Hardware and software developed by RPS gives full pictures of each installation


The verification methods recently developed concentrate on identifying fluid dynamic effects.   In the past this has been ignored but has been shown to be the real cause of metering errors.  This includes not just the inlet pipe but all fittings within 30D of the meter. Take Figure 10 below where two inlet bends in the same plane are followed by a reducer to ensure meter velocity is kept high.  Downstream of the reducer 5D is used in accordance with standards.  It is assumed this meets installation requirements but in reality it does not.  If fittings are located within 5D of each other they interact causing bulk distortion within the flow.  The 5D allowed therefore becomes completely insufficient.

ITA Enews is a newsletter of the Instrumentation Testing Association
A non-profit technical and educational association dedicated to promoting reliability and performance of environmental instrumentation and automation technologies. Learn More About Becoming A Member

Figure 10: Complex inlet pipework example

Verification is not just about assessment; it should also be full documentation and asset assessment.  Frequently installation drawings are not available or may be incorrect.  Producing a full ‘as found’ report as part of the verification process allows users to appreciate the readings from their instruments.  Often fluid dynamic effects causes an instability and meter fluctuations. These are usually removed by long time constant usage without really understanding what the meter is saying.  

28 DAF installation4.JPG


It is important that at least one meter within a works balance is fully assessed and verified. (This is the benchmark). This could be the inlet or the outlet.  The latter is preferable as it directly confirms the DI and can then be used to assess trunk main losses with high confidence.  Sometimes however, accurate in-process meters combined with accurate inlet meters can be used to calculate the works output (DI).  Provided full uncertainties are known for the key meters this method is also acceptable.  Normally it is recommended that the greatest attention should be focused on the output meters.  This is what network managers really need to know accurately.   Consider the simple example below. 


Figure 11: Simple works balance


This shows that for modest values of uncertainty for each of the meters a rather larger works balance uncertainty can be obtained. If we now take the extremes of these values a rather interesting picture emerges.  For the lowest input and the maximum compensation flow the output is (236.4 – 12.6) = 223.8 l/s (within limits of output uncertainty).  For the highest input and lowest compensation flow the output is calculated as (259 – 11.4) = 247.6 l/s.  This is higher than the maximum allowable under the assessment of the outlet meter alone. The questions that now arise therefore are:


· Which meter has an additional bias (from say a telemetry system)?

· Is the input reading high due to age?

· Is the output uncertainty for DI correct?

· Could DI reading be low due to swirl?

· Is the flow uncertainty on the small compensation meter too low?


Only a full verification will allow these to be assessed and fully answered. Imagine the complexity for a larger waterworks where the input source may be from 3 or 4 main meters and the output goes to several zones, measured by multiple meters in parallel. Such a real case in the UK found the works balance was never known and work is currently on-going to assess the actual works balance and therefore quantify the DI. 


It is clear that current industry practices do not fully treat installed measurement precisely.  Single path ultrasonic meters or meters installed where the velocity is low can give rise to large uncertainties.  Waterworks frequently use single or multiple path devices as works output or a system balance meters.   Consider again Figure 5 this time modified to show the meter paths.   There are two parts to the figure the first being a single path meter mounted horizontally (A) or vertically (B).  In the second case a two-path meter is examined, with the beams equidistant from the centreline on the horizontal (C1 and C2) and in the vertical (D1 and D2).   Again very interesting scenarios can be discussed.     


Take the top diagram first.  Calibration in the laboratory under controlled conditions gives the same result whether the beam is horizontal or vertical.  In the abnormal flow case (R) the horizontal installation reads about right but the vertical installation reads low.  If the swirl was at the top instead of at the left side the reverse case would be obtained. (Horizontal reads low and vertical about right).  If the position of the swirl is not known then what is the confidence in the data?


Figure 12a: Single path measurements


Figure 12b: Two-path measurements

Now take the two path case. Again in the laboratory the readings are fine and the meter factors are set.  In the field things really could go awry.  In the horizontal case both beams read slightly low as they cross longer regions of lower velocity so the output would under-record the flow. If the beams are placed vertically then beam D1’ reads high compared to the reference case (D1) and beam D2’ reads very low compared to the reference case (D2).  The overall effect is the meter reads low. 


Literature review shows this type of behaviour is obtained in the majority of cases.  Two-path meters read low most of the time in abnormal profile and swirl situations, but with on average lower errors than single path meters in the same situations.  Magnetic meters too are affected but to a lesser degree. Insertion meters if used on a single plane can give very large uncertainty. The purchase cost of this type of measurement may be low but the uncertainty could be very high. 


It is therefore of paramount importance that the fluid effects are known or are evaluated through complete verification otherwise the works balance is an unknown.  Just because inlet matches outlet does not mean the output flows (DI) are correct.  This has been the case in many UK waterworks recently evaluated.   Without a full verification we come back to our paper title – Do you really know your flowrates?


This paper questions the current ‘best practice’ approach used to assess water network losses in many countries.  Often meter installations are poor, maintenance is patchy and full uncertainty analysis is not undertaken. An uncertainty analysis that is not based on fluid dynamic principles will mislead Water Companies into thinking their balances are accurate. In most of the waterworks investigated recently in the UK, this new approach allows full and accurate quantification of the works balance.


We gratefully acknowledge support from our many customers and field technicians who contributed to the methodology developed in this paper. The Authors would like to thank RPS for allowing them to publish this paper.


Furness, R. 2008a.  The Cost of Measurement Uncertainty: Plenary paper WE-10: 5th PCIC Conference: Electrical and Instrumentation applications: Weimar Germany June 08.


Furness, R and Spitzer D: 2008b.  New Aspects of Flow Measurement Errors in Large Diameter Pipes.  AWWA Journal Vol 100 No. 7, 54-59.


Hutton, S.P: 1945.  The Prediction of Venturi coefficients and their variation with age and  roughness. Proc. Inst. Civil. Eng. London paper 5932


Lambert, A, Myers, S and Trow S: 1998. Managing Water leakage – Economic and technical issues:  Financial Times Business Report ISBN 1 84083 011 5.


Phair, D: 1997.  Errors Occur Often in Municipal Metering. INTECH Sep. 1997 pages 54-57:  ISA, Research Triangle Park, NC USA


Stauss, T (Ed): 2004. Flow Handbook pages 251-270. ISBN 3 9520220 4 7: Endress + Hauser AG, Reinach Switzerland March 2004


UK Managing Leakage. UK Water Industry Managing Leakage, Using Night Flow Data (Report F). ISBN: 1 898920 11 7


Arthur Arscott, RPS Regional Director for Water has a unique blend of skills and experience spanning more than 45 years and dedicated to water distribution. Arthur’s early work began as a distribution inspector with subsequent roles in supervision and distribution management.


For the past 15 years he has focused on leakage control - including regional responsibility for South West Water’s highly successful leakage initiative. He is an Incorporated Engineer, a Fellow of the Institution of Water Officers and, for the past ten years, has been a Director of RPS Water.


Arthur Arscott 
T: +44 (0) 1392 677 333 


Arthur Arscott

Regional Director

RPS Water

Devon UK

Photo courtesy of RPS Water.

Dr Richard Furness, a graduate in both chemical and mechanical engineering, is internationally known in the worlds of flow measurement and water treatment.  He is a UK Chartered Engineer (PE in the USA)  and was awarded his ISA Fellowship in the USA for contributions to measurement uncertainty and metering. He has been an ITA Member since 2007 and has run seminars and courses in Flow Measurement for the ITA in several US locations.    He is consultant to a number of organizations, including UNDP Adviser to the Government of India.  He runs his own technology business from Gloucestershire UK and is also founder of railway posters in the UK.  



Dr. Richard Furness, CEng, ISA Fellow

T: +44 1452 886 560


Dr. Richard Furness, CEng ISA Fellow

Chief Flow Technologist

Gloucester UK

Photo courtesy of Richard Furness.