Non-contact temperature measurement, also known as pyrometry, is viewed with scepticism by many practitioners of temperature measurement. The manufacturers' technical data document that pyrometers are very accurate and precise measuring devices. In addition to selecting the right pyrometer for the application, it is particularly important to consider the material properties and environmental influences on site.
Measuring errors can be avoided through correct use. The most common causes of errors and ways to minimize them are explained below.
Metrological errors in non-contact temperature measurement applications
Introduction
Emissivity
Pyrometers measure the heat radiation emitted by an object. The infrared radiation emitted by the object depends on its material and surface properties. This radiation property is described by the emissivity ε. For exact temperature measurement, the emissivity must be set on the device. An incorrectly set emissivity can cause considerable errors. Figure 1 shows the temperature deviation (ΔT) for three measured values depending on the wavelength if an emissivity of 80 % is set on the device instead of an emissivity of 90 %. This error increases with longer measuring wavelengths or rising temperatures. The shortest possible wavelength range available for the desired measuring range should therefore be selected.
Especially when measuring metal surfaces with unknown or strongly fluctuating emissivity, the measuring error is considerably reduced by selecting a shorter measuring wavelength. The emissivity of metals increases with shorter wavelengths. At the same time, the influence of errors is lower if the emissivity is incorrectly set.
Especially when measuring metal surfaces with unknown or strongly fluctuating emissivity, the measuring error is considerably reduced by selecting a shorter measuring wavelength. The emissivity of metals increases with shorter wavelengths. At the same time, the influence of errors is lower if the emissivity is incorrectly set.

Fig. 1 Measuring error as a function of wavelength and temperature with 10 % deviation of the emissivity (ε device = 0.8 and ε real = 0.9)
Transmission loss
Optimum conditions apply when the pyrometer has a clear field of view of the object. If media such as dust, gases, smoke, protective windows or opaque materials are in the beam path of the pyrometer, these cause a reduction in the temperature radiation of the object.
If the transmission losses are known, e.g. when measuring through a protective glass (τ=0.95), these can be compensated for by adjusting the emissivity on the device.
εdevice = εobject - τbeam path
εdevice = emissivity to be set on the device
εobject = emissivity of the object
τbeam path = transmittance of the objects in the beam path
If the transmission losses are known, e.g. when measuring through a protective glass (τ=0.95), these can be compensated for by adjusting the emissivity on the device.
εdevice = εobject - τbeam path
εdevice = emissivity to be set on the device
εobject = emissivity of the object
τbeam path = transmittance of the objects in the beam path

Fig. 2 Composition of the radiation received by the pyrometer.
It is more problematic if dust, oil or vaporised materials accumulate on lenses or protective windows over time. The pyrometer then measures a lower temperature with increasing dirt. Regular cleaning of the lenses is therefore necessary. Purging units extend the cleaning cycle. Recently, pyrometers with an integrated contamination level indicator have also become available on the market. An alarm signal is generated if the lens is dirty.
Background / external radiation
The radiation powerΦΣ hitting the pyrometer's detector is decisive for the displayed object temperature.
According to the following formula, in addition to the emission component of the measuring object, it includes a background radiation component consisting of the reflection and transmission component of the ambient radiation.
ΦΣ = Φε + Φτ + Φρ
ε = emissivity of the measuring surface
τ = transmittance of the measuring object
ρ = reflectance of the measuring surface
The error influence of the background radiation is reduced the greater the emissivity of the object and the greater the object temperature compared to the ambient temperature. This influence is problematic, for example, when using pyrometers at the outlet of continuous furnaces. The measuring error can be reduced if the alignment of the lens prevents reflection of heat radiation from the furnace on the surface of the measuring object. Radiation sources in the infrared range, such as incandescent lamps, radiant heaters or lasers, sometimes cause strong infrared radiation, which is underestimated in practice.
Devices with blocking filters are available especially for laser applications in order to prevent the influence of high-energy laser radiation on the very low infrared radiation.
According to the following formula, in addition to the emission component of the measuring object, it includes a background radiation component consisting of the reflection and transmission component of the ambient radiation.
ΦΣ = Φε + Φτ + Φρ
ε = emissivity of the measuring surface
τ = transmittance of the measuring object
ρ = reflectance of the measuring surface
The error influence of the background radiation is reduced the greater the emissivity of the object and the greater the object temperature compared to the ambient temperature. This influence is problematic, for example, when using pyrometers at the outlet of continuous furnaces. The measuring error can be reduced if the alignment of the lens prevents reflection of heat radiation from the furnace on the surface of the measuring object. Radiation sources in the infrared range, such as incandescent lamps, radiant heaters or lasers, sometimes cause strong infrared radiation, which is underestimated in practice.
Devices with blocking filters are available especially for laser applications in order to prevent the influence of high-energy laser radiation on the very low infrared radiation.
Nothing beats a good lens
Imaging errors in the lens, scattered light and reflection from optical components and housing parts as well as diffraction due to the wave nature of the light result in some of the detected radiation reaching the sensor outside the specified measuring field. The lens receive part of the radiation outside the measuring field. This influence of the lens is known as the "size of source effect". This influence can be minimized by careful correction of optical imaging errors, the use of anti-reflective optical components and the avoidance of scattered light and reflections in the device. High-quality lenses reduce these error influences. The "size of source effect" is at its smallest in the focus of the lens. In pyrometers with focusable lenses, this effect can therefore be significantly reduced if the measuring distance is set correctly.
The optical error increases with the wavelength for physical reasons. Therefore, even more effort is required for optical error correction in the case of long-wave measuring devices and thus devices for low measuring ranges. With cheaper pyrometers that measure from room temperature, this has a negative effect in that the measured value displayed is very dependent on the selected measuring distance.
If the object is significantly larger than the pyrometer's measuring spot and the surface is at almost the same temperature level, this effect can be almost neglected. Otherwise, the error can be reduced by using a device with focusable lens and exact alignment to the object. A pilot light, a through-the-lens sighting or an integrated video camera are recommended for exact alignment of the pyrometer.
The optical error increases with the wavelength for physical reasons. Therefore, even more effort is required for optical error correction in the case of long-wave measuring devices and thus devices for low measuring ranges. With cheaper pyrometers that measure from room temperature, this has a negative effect in that the measured value displayed is very dependent on the selected measuring distance.
If the object is significantly larger than the pyrometer's measuring spot and the surface is at almost the same temperature level, this effect can be almost neglected. Otherwise, the error can be reduced by using a device with focusable lens and exact alignment to the object. A pilot light, a through-the-lens sighting or an integrated video camera are recommended for exact alignment of the pyrometer.
Two-colour pyrometers
With a two-colour pyrometer, the ratio of the radiation densities of two different spectral ranges is analyzed. Simplified, the following formula applies to the measured temperature with the two centre wavelengths λ1 and λ2.
1 ÷ TM = (1 ÷ TW) + ((λ1 · λ2) ÷ (C2 · (λ1 - λ2))) · (ln {ε1 ÷ ε2})
TM = emissivity of the measuring surface
TW = transmittance of the measuring object
C2 = reflectance of the measuring surface
1 ÷ TM = (1 ÷ TW) + ((λ1 · λ2) ÷ (C2 · (λ1 - λ2))) · (ln {ε1 ÷ ε2})
TM = emissivity of the measuring surface
TW = transmittance of the measuring object
C2 = reflectance of the measuring surface
If the emissivities ε1 and ε2 are the same for both wavelengths, the measured temperature corresponds to the object temperature. A two-colour pyrometer therefore measures independently of the emissivity of the surface, provided the emissivities ε1 and ε2 are identical. In theory, two-colour pyrometers are recommended if the emissivity of the measuring object fluctuates. In practice, however, this is dependent on the respective application and rarely applies. Due to the ratio formation, the measuring error of a two-colour pyrometer can be considerably greater than that of an one-colour pyrometer if the emissivities of the two measuring wavelengths fluctuate and differ. Metals in particular, and especially non-ferrous metals, exhibit a wavelength-dependent change in emissivity.
Transmission losses such as dust, vapour or smoke, on the other hand, often cause a homogeneous attenuation of the radiation intensity. In comparison to one-colour pyrometers, the measured value of two-colour pyrometers remains constant under these conditions.
Transmission losses such as dust, vapour or smoke, on the other hand, often cause a homogeneous attenuation of the radiation intensity. In comparison to one-colour pyrometers, the measured value of two-colour pyrometers remains constant under these conditions.

Fig. 3 Avoid measuring errors due to reflected background radiation by correctly aligning the pyrometer.
Innovative two-colour pyrometers enable the simultaneous measurement and calculation of the temperature at both one-colour wavelengths and the two-colour temperature. This makes it possible to decide during commissioning whether the measurement with a one-colour or with a two-colour pyrometer provides more reproducible and accurate measured values for the entire measuring range.

Fig. 4 Recording of both one-colour and two-colour temperatures with the CellaView software.