The importance of calibrated, traceable instruments for measurement.

Measurement is an aspect of life that most of us take for granted – when buying a litre of wine, whether in Athens, or Brussels or Copenhagen we expect to get 1 litre. In fact the reason we get the same measure is not due to chance – it’s due to the standardised national measurement systems of the EU.

The need for accurate measurement has been with us since the beginnings of history. It has been suggested that in the Stone Age there was a standard unit of measurement – the megalithic yard – which was used to design prehistoric monuments such as Stonehenge 1 . Certainly by the Bronze Age standardised length, mass and time measurements were in use in the Indus Valley.

In more recent times, this need for standard weights and measures has been enshrined in law by some of the most fundamental acts in history 2 – e.g. the UK’s Magna Carta (1215) and the US Constitution (1787). Since 1875 the Convention du Mètre has provided mechanisms for governments to co-operate on measurement standards. One reason that metrology is so important is that it drives and underpins international trade, legislation and regulations. It is estimated that in Europe a cost equivalent of 2%-7% of GDP is concerned with measurement 3 .

Measurement standards also engender trust – they ensure that the calibration certificates for equipment manufactured in one country are also accepted in the country of receipt. Quality Management systems address this need for instruments to be calibrated, but the question arises calibrated to what? National and international standards sit at the top of the traceability chain providing a reference point for all calibrated measurements.

Measurement also underpins legislation e.g. ensuring that substances can be monitored. However, the effect of legislation can cascade in unexpected directions and have measurement consequences outside the intended area. The EU recently proposed a strategy that bans the use of mercury 4 in certain applications. Mercury, especially in its vapour form, is highly toxic and accumulates in food chains. It is clear that methods must exist to measure mercury content.

However less obvious is the need to measure and calibrate tympanic (digital ear) thermometers. Although mercury will not be banned in specialist applications, the general public will have to use non-mercury based thermometers. Ear thermometers are a readily available substitute. However they are hard to use correctly and without calibration can give an incorrect answer 5 leading to false reassurance or misdiagnosis. Legislation designed to protect the public in one way, could be increasing their exposure to risk in another due to inappropriate measurements.

Measurement standards

The Convention du Mètre gives authority to organisations such as the General Conference on Weights and Measures (CGPM) and the International Bureau of Weights and Measures (BIPM). Central to these structures and agreements are individual countries’ National Measurement Institutes (NMIs). The National Physical Laboratory (NPL) is the UK national metrology Institute, which develops and maintains the national measurement standards for the UK, and crucially carries out extensive R&D to keep the UK as a centre of excellence in measurement.

These standards form the basis of measurement infrastructure in the UK ensuring accuracy, consistency and repeatability throughout traceable measurements. Collaboration with other eminent NMIs in Europe (and worldwide) is increasingly important, as challenges in air, water and environment measurement do not respect national borders. NPL is a key player within EUROMET, the collaboration of the European NMIs (e.g. PTB in Germany or LNE-INM in Paris), and is leading an initiative to facilitate collaborative R&D amongst the NMIs as they gear up to address the “grand challenges” that will shape the future of our economies and our way of life.

Uncertainty, precision and traceability

It is important to know the accuracy of any measurement instrument i.e. does the measurement truly represent what is being measured? As illustrated in figure two, it is possible to make an accurate but imprecise measurement or be precise but inaccurate. A systematic error in the measurement (caused by a number of factors that may include drift or malfunctions) could lead to a difference between the true value and the measurement value.

To minimise all such errors, the measurement method should follow standardised procedures (usually specified through a quality system such as ISO/IEC 9001:2000 7 ) and the instrument should be calibrated periodically. Figure one illustrates the UK traceability chain for any traceable physical measurement. Traceability can be obtained through accredited calibration laboratories to the national standard held at an NPL (e.g. the International Temperature Scale for temperature).

Calibration is the process of attributing the traceable measurement to a device and must be preformed periodically to ensure the device is continuing to operate properly. Accreditation is an objective means of reviewing a calibration procedure. In the UK a third party (the United Kingdom Accreditation Service (UKAS)) reviews and assesses calibration procedures, records, equipment and staff who carry out calibrations to ensure that any test or calibration laboratory meets the requirements of the international standard and is performing validated traceable measurement. The accreditation standard ISO/IEC 17025:2000 8 covers the above requisites.

Temperature and thermophysical properties measurement

The International Temperature Scale of 1990 (ITS90) is the standard for temperature measurement. For a temperature measurement to be trusted it should be made by an instrument whose calibration can be traced back to the scale. It extends from 0.65K upwards to the highest temperature practically measurable in terms of radiation law (normally ~3000 °C). It consists of a series of reference points where the temperature is defined e.g. the triple point of water is 0.01 °C or the freezing point of Silver at 961.78 °C. Beyond this highest defined temperature point, the temperature is extrapolated using pyrometry.

Below 0.65K a provisional scale provides the temperature. This means that although it is possible to buy solid-state cryogenic refrigerators that can cool to less than 0.65K, it is not possible to say exactly what temperature has been reached.

The measurement of temperature (and thermal properties) has impact in a huge range of fields. Accurate temperature and thermophysical measurements reduce energy consumption, cut product waste and increase efficiency. Good temperature control is often essential to ensuring the consistency and quality of a product.

For example, to reduce the fuel consumption of passenger aircraft, jet engines are required to operate at ever-higher temperatures. The turbine blades must undergo a heat treatment at over 1310 °C for several hours to provide the necessary metallurgical qualities. This treatment requires temperature tolerances +2/-3 °C which is the lowest uncertainty currently available from thermocouple manufacturers. The heat treatment is therefore limited by available uncertainties, and additional destructive testing is required to verify the strength of the blades after treatment.

Current best uncertainties at NMIs are +/- 1 °C. NPL is currently working on a project to transfer these uncertainties into industry. Successful lowering of the uncertainties should result in greater effectiveness in heat treatment processes, better process control and lower energy consumption.

Temperature control is also vital to meeting the numerous stringent regulatory requirements of many sectors e.g. food, health, safety and pollution control. Temperature measurement, although not often the direct subject of legislation, is often one of the measures that enables and underpins the legislative requirements. With so many different applications, there are many different methods of measuring temperature: contact methods such as thermometers, thermocouples; non-contact methods such as pyrometry (measuring the radiation emitted from an object) and novel methods such as acoustic measurements of flame temperature.

Humidity measurements

Like temperature, humidity measurements 9 impact across a large range of sectors including energy supply, food processing, electronics and environmental testing. In the UK there are 15 UKAS accredited laboratories offering traceability back to the UK’s national standards held at NPL. Humidity is a relatively difficult measurement because it depends on so many factors. It is particularly important to consider temperature effects when measuring humidity. Generally the hotter the air (or other gas), the more water vapour it can hold. At any particular temperature the point where the air contains its full capacity of water vapour is called its saturation point. Various terms are used to describe humidity:

  • Relative Humidity: how fully saturated the air is compared to the saturation point (at a given temperature). Often referred to as ‘rh’, it is expressed in units of percent
  • Absolute Humidity: the mass of water vapour present in unit volume of moist air at a given temperature and pressure. (When minute qualities of water – less than 1 ppmv – are present in a gas, this is sometimes known as Trace Moisture content)
  • Dew-point: the temperature at which condensation forms on cooling a gas (in effect the point where the air becomes saturated with water). Note that the liquid is not necessarily water, nor the gas air

Again, as with temperature, there are a variety of methods available to measure humidity. The method selected depends on the range, the temperature, the uncertainty required and the application – but it is worth doing: in recent work at a cheese manufacturer, accurate humidity measurements have enabled annual savings of £15,000 and an increase in product yield.

Measuring the energy performance of buildings

Under the Kyoto agreement, the UK agreed to cut their CO 2 emissions by 12.5% of 1990 levels by 2012. It is estimated that 40% of energy consumption (linked to CO 2 emissions) is spent in the space heating and cooling of buildings. The wholesale cost of gas and electricity has risen by over 150% since 2004. Both these drivers should ensure that consumers should understand the benefits of reducing energy usage, but it is often not clear to them the substantial savings that could be made.

To assist in the process the EU issued Directive 2002/91/EU on the energy performance in buildings to provide a ‘legal instrument … to lay down more concrete actions with a view to achieving the great unrealised potential for energy savings’ 10 . The directive is designed to be complementary to the directive 93/76/EEC to limit CO 2 emissions. The member states had until 4th January 2006 to pass appropriate laws, and a further three years in which to train appropriate inspectors if required. In the UK, this directive has been passed into law through ‘The Building and Approved Inspectors (Amendment) regulations 2006’. Two of the requirements under the new regulations are:

  • Energy efficiency standards for both new and renovated buildings will have to be improved to a value determined by the Secretary of State in the form of CO 2 emissions rates
  • An ‘Energy Performance Certificate’ for the building when it is sold or leased

The regulations allow for ‘a methodology of the calculation of the energy performance of buildings’. This is a model that calculates the annual energy used for a proposed or existing building and compares it with a ‘notional’ building. The directive is not prescriptive in specifying which calculation should be performed – it only identifies what aspects need to be considered. However even if all the models across the EU were the same, it is only as good as the physical data used as inputs.

And here there is a problem as Tony Phillips of the British Board of Agrément (specialists in measuring and assessing the performance of building products) describes ‘there is not only an absence of physical data but no commonly agreed methodology for obtaining the data’.

Very few organisations have the necessary facilities capable of measuring all the required U-values (a measure of the amount of heat lost through 1 m 2 of the material for every degree difference in temperature either side of the material) to traceable standards. In the UK only NPL and the British Board of Agrément possess validated hot boxes. NPL is heavily involved in ensuring that the measurements for the building calculations are valid and conform to the international standards (e.g. the EU product standard for industrial doors (BS EN ISO 12567-1) or for insulated roof and wall panels (BS EN ISO14509:2004).

It is only through accurate, traceable thermal property measurements such as those performed at NPL that the calculations required by Energy Performance of Buildings Directive will have meaning. If the calculations are not accurate then neither is the Energy Performance certificate and the end goal of reducing energy use and carbon dioxide emissions will be harder to achieve.

Measuring trace moisture in industrial gases

Trace gas measurements are being increasingly required across a wide range of industrial sectors, including petrochemical processing and power generation. Many of these measurements are made in matrices other than air or nitrogen. HCl is a process gas with increasing applications in a number of industrial settings particularly in the semiconductor manufacturing industry. Moisture is a contaminant both in its own right and as a source of oxygen molecules, which can lead to damaging oxidation in a wide range of sensitive processes. The presence of even very low levels of moisture in HCl gas is therefore of increasing concern, and the nature of the consequences and interactions is not fully understood. 11

Measuring the temperature of the sun

One of the burning issues of modern times has been the issue of climate change: is it happening; is it caused by mankind’s actions; if so what effect will it have? There are many influential people who continue to deny that climate change is a problem. Why is this? One of the main reasons is that measurements made have too great an uncertainty to be definitive about the result. Thus acceptance of climate change becomes a matter of belief rather than absolute scientific ‘fact’. In order to change this it is necessary to measure the temperature relating to climate change rather better than it is done currently.

Earth observation from satellites is extremely useful. Many applications do not need high certainty because it is the difference in a spatial scene or over short times that is important e.g. identifying vegetation type or looking at the spread of pollution (e.g. industrial effluent). However if measurements are required quantitatively or over a long time period then measurement accuracy becomes critical. To determine how warm it will be in 2100, various models have been created. Each model has a range of results depending on the accuracy of the underlying measurements.

The mean temperature rise is currently predicted to be 4.5 ºC – see ‘Monitoring and Forecasting the Weather’ in the June issue – but there is a large uncertainty associated with this value. The models need information about the sun – what is the solar irradiance and how does it interact with the Earth System – i.e. what heat should be input into the model. The problem with this measurement is that it comes from data from a number of satellites. These have been operating at different times and since their measurements are not strictly traceable they cannot be compared.

Those measuring input solar irradiance have a variation in the absolute level of their measurements of >0.7% (although an individual satellites will typically see ~ 0.1% over an 11 year cycle). This is a significant amount – around 1650, the River Thames frequently froze and the temperature was 2 degrees lower than it is now. This corresponds to the Sun cooling by 0.3% – half as much as the variation in our current instruments.

There is an additional problem. When the sunlight reaches the earth, using the satellites its possible to observe the flow, absorption and reflection of incoming radiation. Accurate measurements are vital to the input data to the climate change models, but when the measurements are made there is a discrepancy – approximately 20% more energy is absorbed in the atmosphere than the models predict. This amount has up to ten times the heating effect than from greenhouse gases. It is thought to be due to water vapour (i.e. humidity) and aerosol pollutants in the atmosphere.

NPL is trying to develop solutions to these problems by improving the measurement techniques and the calibration and traceability of the instruments. For example in 1970 NPL introduced a new type of radiometer that is 100 times more accurate than the ones currently used in space. They are used to calibrate the satellite instruments on the ground but once in space it is not possible to recalibrate. However NPL is proposing a solution – a satellite called TRUTHS (Traceable Radiometry Underpinning Terrestrial and Helios Studies).

It will reduce all the uncertainties by essentially providing regular calibrations in space by benchmarking reference targets for other satellites. It will also enable all previous observations to be linked together as well as those obtained from other means such as balloon flights. NPL is also working on atmospheric measurements through LIDAR to help understand the discrepancies in the absorbed sunlight.


Measurement of temperature and humidity is complex and vitally important to a huge range of industries, healthcare and society. But what is equally important is that the measurements are made with calibrated, traceable instruments. Only if this is done, is it possible to understand individual scenarios and underpin any legislation.


1 Thom, Alexander, ‘Megalithic Sites in Britain” 1967

2 Redgrave, F. ‘Metrology – who benefits and why should they care’, NCSL International Workshop and Symposium, 2005

3 Williams, G. ‘The Assessment of the Economic role of Measurements and testing in Modern Society’, EU DG Research, Contract G6MA-2000-20002, July 2002

4 Commission of the European Communities (2005) Community Strategy Concerning Mercury, COM(2005) 20

5 Mackechnie, C. and Simpson, R. ‘Traceable calibration for blood pressure monitoring and ear temperature measurement’ Accepted for publication Nursing Standard

6 MHRA (2003) Medical Device Alert MDA/2003/010, MHRA

7 ISO/IEC 9001:2000, Quality Management systems, requirements, 2000

8 ISO/IEC 17025:2000 General requirements for the competence of testing and calibration laboratories, 2000

9 ‘A Guide to the Measurement of Humidity’, The Institute of Measurement and Control and NPL, 1996 ISBM 0-904457-24-9

10 Directive 2002/91/EU ‘On the energy performance of buildings’, OJ of the EC, 4th January 2003

11 ‘Humidity In Industrial Gases’ meeting to be held at NPL, 11th October 2006. Please visit for details

Published: 01st Sep 2006 in AWE International