Young adult subjects who occupied a well-ventilated space with low background of level of ozone achieved via carbon-filtration could detect ozone odor at 7 ppb, lower than expected from archival compilations. The outcome was not inconsistent, however, with some observations of recognition, beyond mere detection, at about 15-20 ppb. Individual differences in sensitivity lay at or just below an order of magnitude, rare in olfactory testing and indicative of precision. In a study of d-limonene, subjects again showed high sensitivity and small individual differences. The subjects could detect the odor at 8 and 15 ppb, depending upon whether they occupied a space with or without carbon filtration, respectively. The results argue for use of carbon filtration to measure sensitivity most stringently, although absence of filtration seems not to incur a large penalty. The protocol used here, with collection of hundreds of judgments in a day, yet with little net exposure of the subject to odorant; with verifiably stable delivery; and with analytical confirmation of level should reduce tolerance for outcomes of large differences among subjects and among studies.
Practical implications: Humans manifested much higher sensitivity to ozone and D-limonene than commonly thought, a pattern revealing itself more broadly in olfactory studies as testing improves and analytical confirmation of delivery becomes more common. Published databases, with errors of +/-1000%, often badly underestimate sensitivity and can thereby encourage use of higher concentrations of compounds, particularly VOCs, than relevant in studies of reactive indoor chemistry.