COMPARISON OF OLFACTOMETRY, GAS CHROMATOGRAPHY, AND ELECTRONIC NOSE TECHNOLOGY FOR MEASUREMENT OF INDOOR AIR FROM SWINE FACILITIES

Indoor air from swine finishing facilities was analyzed by olfactometry, gas chromatography, and an electronicnose. Odorous air samples were collected from two feeding rooms at an Iowa State University facility. Six finishing pigs werehoused in each room. Tenliter room air samples, collected in Tedlar bags, were analyzed using the ACSCENT InternationalOlfactometer (St. Croix Sensory, Stillwater, Minn.). Dynamic dilution triangular forcedchoice olfactometry was the methodemployed. Oneliter air samples, also collected in Tedlar bags, were analyzed using an AromaScan A32S electronic nose(Osmetech, Crewe, U.K.) containing an array of 32 conductingpolymer sensors. Solid phase microextraction fibers(Supelco, Bellefonte, Pa.) were used to adsorb compounds in the air to be analyzed by gas chromatography/mass spectrometry(GC/MS). Sixteen compounds, primarily nonsulfur protein metabolites, were developed into a quantification standard forGC/MS based on the frequency of occurrence of these compounds in the collected samples. Using these compounds, anequation was generated to predict odor dilution threshold. Poor prediction capability (R2 < 0.3) indicates that additionalanalytes require consideration. Electronic nose evaluation of room air samples was not strongly correlated to olfactometrymeasures (r < 0.2). However, the equation developed from the GC/MS analyses was capable of predicting the electronic noseresponse to air samples (R2 > 0.8). The results suggest that human panelist responses may be based on detection of compoundsthat were not included in GC/MS quantification procedures and are not well detected by this electronic nose. Variation inresponse within and among human panelists likely accounts for some of the discrepancy between human assessment andchemical and instrumental methods of odor evaluation.