Beyond Detectability

ABSTRACT

Part of the input to a dispersion model consists of a source odor concentration that typically is determined by a panel of six to eight observers using a forced choice dynamic triangle olfactometer. A series of dilutions of the sampled odor are presented and the dilution at which one-half of the panel can detect odor is considered to be the source odor concentration (often described as ED50 or as Dilutions-to-Threshold [D/T]). The output from the dispersion model generates a predicted ambient odor concentration (expressed as D/T or as odor units). An ambient odor concentration (D/T) of 1 means that 50% of the panel will detect odor. It is not clear what D/T values other than 1 mean. Thus, only limited information is given on the detectability (threshold) dimension of odor.

In this paper, a procedure is outlined that can be used to obtain information on all four dimensions of odor and still keep the simplicity and usefulness of the source odor concentration/ambient odor concentration concept that the dispersion modeler needs. The panel members are asked to report for each dilution presented the four dimensions of odor-detectability (as a “yes” or “no” odor response), intensity (butanol or category scale), quality (odor descriptors), and the he-donic tone. The paper illustrates how the procedure can be used with either inexperienced or experienced panels who provided information on three (detectabil-ity, intensity, and quality) of the four dimensions of odor. The procedure is applied to odors from two wastewater treatment plants. Essentially a dilution profile of the sample is prepared. For a given D/T, the procedure can provide an estimate of the percent of the panel that will detect odor, the average odor intensity, and the odor quality.