Sensory evaluation method of building materials for labelling purposes

It has been shown that odour is an important factor of material emissions and the odour of a product can very seldom be predicted by the chemical emissions. A full size chamber of 5 m³ equipped with a sniffing hood was built to enable both chemical measurements and sensory assessment to be performed using the test conditions of chemical testing. To compare the performance characteristics of the new 5-m³ chamber with the conventional chambers six different materials were tested both in 1-m³ chamber, sensory assessment chamber and in the 5-m³ chamber. The sensory assessment chamber used for the comparison was a 100-litre aluminium chamber, which was operated according to the principles given in Nordtest method NT Build 482 (diffusor). Statistical studies were performed to verify the panel size required for sensory tests for labelling purposes. Chemical and sensory test results obtained with different chambers were in good agreement. Comparing the results with the Finnish building material classification the chamber size would not have affected the labelling decision the materials. Statistical calculations presented here show that for labelling purposes reliable sensory acceptability evaluations can be performed using a relative small naive, untrained panel. When accepting uncertainty of 0.4 on the acceptability scale and a symmetric risk of 10 %, a panel of 5 members can be used. In those cases when the mean acceptability is near the labelling limit value, a bigger panel with 15 members is needed to ensure uncertainty of 0.2 acceptability units and 10 % risk level of making a wrong decision.