A dynamic olfactometer is a gas diluting
apparatus and also an interface between a panel of human observers
and an odorous gas sample diluted at various concentrations. Olfactometry
requires a very high standard of testing conditions. These include
an odour-free testing environment, a highly accurate and repeatable
olfactometer and effective panelist management.
Odour-free testing environment:
An odour-free testing environment is an important element in the
olfactometry testing process. An odour-free room provides not
only a relaxed resting environment for the comfort of panelists,
but also eliminates background odour that may lead to olfactory
adaptation and fatigue, which would otherwise affect the sense
of smell of panelists and result in a failure to detect odour
at low concentrations. The test room air should be filtered using
an activated carbon filter with a minimum air exchange rate of
at least 12 times per hour.
Olfactometer calibration:
The olfactometer must be calibrated against a tracer gas to check
that the dilution setting of the olfactometer meets repeatability
and stability criteria. It is not sufficient to check the dilution
setting simply against gas flow rate. Firstly, the olfactometer
must be able to demonstrate its stability within 5% at each dilution
level. This ensures that all panelists receive the same level
of odour concentration. Secondly, the olfactometer must be able
to repeat each dilution level within 20% of the setting value.
This ensures that the dilution steps will be evenly distributed
across the testing range. Further details are provided in Dutch
standard NVN 2820 and the draft CEN/TC264 standard.
Panelist management: Panelists
should be trained and screened using reference air incorporating
certified n-butanol at a concentration of 60 ppm using the same
procedure as used for environmental samples. Means and standard
deviation are calculated for 12 logarithms of individual thresholds
(break points). The criterion is that means should be in the range
of 3.00 to 4.38 with a standard deviation ofless than 0.916.