Problems of measurement with thermal imagers in determination of energy efficiency of buildings

The paper describes problems in the use of thermal imagers for determination of energy efficiency of buildings. Thermal imagers are commonly used in technical diagnostics and measurement of energy efficiency of buildings. A thermal imager captures an infrared (IR) image with an array of detectors (microbolometers). Analysis of an IR image is made with an associated software, in which parameters such as emissivity, distance, temperature and humidity are defined. According to recent analyses, only some 10 % of thermal imagers are accurate within the set manufacturers specifications. Therefore, it is important to use calibrated thermal imagers, when absolute temperature values are important. There are less problems when measuring temperature differences, but not all thermal imagers are capable of measuring them accurately. In their practical use, the following effects have to be taken into account: operating conditions (temperature, relative humidity), thermal capacitance, night sky, influence of athmosphere and distance, angle of measurement, reflected radiation from the surrounding, wind, etc. In our practical examples we show measurement of house and snow with two different thermal imagers (low cost and expensive) under the same measurement conditions. Figures 1A, 1B and 3 (A, B, C, D) were made with a low-cost thermal imager (resolution of the IR detector 160x120), while Figures 2A, 2B and 4 (A, B, C) were made with an expensive thermal imager (resolution of the IR detector 320x240). Both thermal imagers were calibrated in our laboratory under laboratory conditions and showed good agreement with the manufacturer specifications regarding accuracy (+/- 2 degrees C or 2 % of reading for the low-cost thermal imager and +/- 1 degrees C or 1 % of reading for the expensive thermal imager. Our analysis of the measured results shows good performance of the expensive thermal imager not only in measurement of temperature differences (Table 1B) but also in measurement of the absolute temperature (Table 1A). Results obtained with the low-cost thermal imager are completely different. Here all the measurement values are questionnable and the majority of them are physically quite impossible. It is obvious that the low-cost thermal imager is sensitive to changing operating conditions, which was especially evident when measuring snow. Before choosing a thermal imager its purpose and required accuracy should be defined. Measurement of the absolute temperature with accuracy better than +/- 1 degrees C is hard to achieve with a thermal imager, even if the manufacturer specifications claim so. Accuracy of a thermal imager is determined by calibration in an accredited laboratory issuing a calibration certificate. As a rule, thermal imager manufacturers do not provide them, at least not for free. Moreover, a calibrated and accurate thermal imager does not necessarily provide accurate measurements. To assure accurate measurements with an accurate thermal imager, a competent and experienced operator is required, too.