The amount of dilution required to achieve odorless air delivers a crude point of departure but should acknowledge the large error and be presented with only one significant figure in the final results. Thresholds for odorants typically vary by several orders of magnitude, especially the ODTC50 . Further, reliable methods may not have been used, and results for odor detection and odor recognition are sometimes mixed up. Within a controlled setting, two approaches to sensory testing of dilutions by panelists are used. One is the Odor Profile Method , which uses sugar solutions for calibration, and the other uses odor disappearance upon sufficient dilution. Both rely on dilution equipment, typically a dynamic dilution olfactometer that delivers sampled air diluted with odorless air to a nose port where the dilution is smelled by the panelist. Concentrations are presented in ascending order to avoid desensitization and anticipation bias. Statistics are then used on the panelists’ results to determine the ODTC50 for the odorant or mixture tested. For a single odorant, OPM is used to assign an odor note and intensity to each dilution. The Weber-Fechner law is applied, meaning that the logarithm of the concentration is taken and then the intensity results are fit to a line through linear regression. Extrapolation to intensity score 1 yields the ODTC50 value. Each odorant can have a vastly different slope. For example, a 200-fold change in concentrations of 1-propanol and n-amyl buterate caused a 15-fold and 0.5- fold change in odor intensity, respectively . For mixtures , the ODTC50 can be determined by forced choice, typically “triangle,” methods using the same dynamic dilution equipment . While inhaling at the nose port,commercial racks the panelist rotates between three choices and then selects which is the diluted sample. A point estimate is generated rather than a curve, and no odor description is gathered.

ASTM Method E679-04 has been used to determine the ODTC50 values for a range of odorants . Such methods have been used in the drinking water industry to set ODTC50 values for methyl tertiary butyl ether and improvements to ASTM Method E679-04 have been suggested for drinking water . In Europe, under EN 13725, the final dataset typically only includes the data for the four or more panelists whose results are the most consistent with the overall panel’s geometric mean value. Also, panelists may be presented with 2 samples instead of 3. The dilution result is called “European odour unit” , which is defined as “the amount of odorants that, when evaporated into 1 m3 of neutral gas at standard conditions, elicits a physiological response from a panel equivalent to that elicited by 1 European reference odour mass [123 μg n-butanol] evaporated in 1 m3 of neutral gas at standard conditions” . Thus, the European approach accounts for the variation in detection thresholds of the panelists. Despite these strictures, proficiency tests in 2007 and 2008 found two thirds of the European laboratories claiming to work in accordance to the EN 13725 standard failed to demonstrate compliance with the required performance criteria . Dilution equipment can also be used to evaluate the odor hedonic tone, as is required in the Netherlands . Panelists express the degree of pleasantness using a 9-point scale ranging from H -4 to H +4 . The logarithm of the odor concentration is plotted versus the H-value, which approximates a straight line. Using linear regression, the level of dilution required to achieve Dutch regulatory criteria can be estimated. OPM used to determine the ODTC50 goes against the guidance upon which it was built, yet still delivers useful results . Both the FPA and APHA Method 2170 advise against conducting statistical analysis on the intensity scores. The scores are categorical rankings and not an interval scale. In other words, an intensity score of 8 is not necessarily twice as intense as a score of 4. The intensity scale was not designed for 1 to be the ODTC50 value, and historically used the symbol. Surprisingly, such “against the guidance” extrapolation was found to be not off-the-mark from ODTC50 values from other studies , which may be confirmation of the enormous range of such values .

Day-to-day variation by panels is typically within an order of magnitude . Faint environmental odors are sensed but cannot be measured by dynamic dilution olfactometry . Typically, dynamic dilution values are never given for levels lower than 10 OUE/m3 , and generally the lower values start in a range of 50–100 OUE/m3 . The use of human subjects as panelists is strictly controlled . International codes of conduct apply as do reviews by a human subjects review boards to protect participants.Exposure limits are intended to protect the health of populations and arrive at quite different results for workers versus the general public. Workers are considered healthier and less diverse in susceptibility than the general population. They also only spend a portion of their day and week on-the-job. Political pressures, too, may influence worker limits to prevent onerous restrictions to industry . The morbid or lethal outcomes of occupation exposures likely overshadowed concerns about impact on olfactory function. The end result is that worker exposure limits are usually many times higher than those for residential exposures. A survey by National Geographic revealed that factory workers reported poorer senses of smell , although sensory loss is typically gradual and may be confused with aging . Sensory-impaired workers can miss out on warning signals that impact their nutrition and quality of life. Monitoring of workplace air for hazardous chemicals cannot be replaced by sensory cues, due to a portion of the workforce having little or no sense of smell , odors can create a workplace nuisance for most employees. Workers, too, are subject to the psychological strain of odor exposure, and setting and occupational exposure limit above the odor threshold may lead to perceived risk and well as physical response . For these reasons, occupational exposure limits consider odor. OELs have long considered irritation . In the United States and Europe, about 40% of the OELs are set to avoid sensory irritation . Sensitization is especially a concern, and OELs should be set low enough to avoid such initial, triggering exposures .

Only three chemicals are regulated by OSHA based on “obnoxious odor” and worker complaints: isopropyl ether, phenyl ether, and vinyl toluene . Critical reviews suggest setting OELs by taking into account both irritation and odor thresholds . For residential exposures, sensory effects are considered in emergency situations but rarely for ongoing, long-term exposures . The latter protect from cancer and non-cancer effects by focusing on system organ toxicity. Whether odor is a health-protective signal varies from odorant to odorant. Carcinogens have especially low exposure limits, so the ODTC50 tends to be well above the residential exposure limit. Benzene and formaldehyde are prime examples of such carcinogens . Because ODTC50 values range over several orders of magnitude, they commonly overlap with exposure limits. Re-visiting the work by Rosenfeld et al. and adding the ODTC50 ranges from AIHA , this overlap became apparent for hydrogen sulfide and ammonia .Exposure assessment involves the measurement of odorant concentrations in the hope the subset measured is responsible for the odor. It also involves the use of human panels to identify odor characteristics . Such categorical measures cannot be used in a quantitative risk assessment; however, they do provide qualitative information. The measurement of the number of dilutions required to remove an odor is plagued by inconsistent results and driven by the final remaining odorant in the dilution . When exposure measurements have been determined for a subset of odorants, the readings can be compared to the exposure limits presented in Section 4.4. This is the risk characterization. For residential exposure, the comparison is with the ODTC50 by a ratio known as the “Odor Activity Value” . Such a comparison, however, carries with it the inherent weakness in the numerator and denominator values in the ratio, undermining some of its usefulness. Specifically, ODTC50 values usually span odors of magnitude, and measured concentrations are typically a snapshot in time. Some studies find the inputs to OAVs too variable and are now pursuing other methods instead . In addition, for ongoing rather than intermittent odor exposures, comparisons with health based exposure limits for residents, such as the USEPA Reference Concentration or California EPA Reference Exposure Level ,greenhouse rolling benches can be performed for non-cancer effects. For carcinogens, the cancer risk estimate involves the inhalation USEPA Unit Risk Factor or equivalent. Note that exposure assessment usually has less uncertainty than other portions of risk assessment . Mixtures are challenging for risk assessment. Combined effects of sensory irritants can be considered additive as a first approximation . The interplay of odorants, however, is often unknown . Hydrogen-sulfide-equivalents are a proposed approach , similar to the dioxin equivalents for the 17 dioxin-like congeners and the carbon dioxide equivalents for various greenhouse gases.As a literature review, where no exposure data were generated, the case-study approach is used to illustrate how odorant concentration data can be used in conjunction with odor threshold information to characterize the health risks.

Although the most significant risks are due to chemical exposures in the workplace , the focus of this paper is on residential exposures to environmental odors, so worker exposure studies were excluded .Not only have OAVs been evaluated for on-site worker exposures , they also have been evaluated for odorants from sewers across Australia over a period of 3.5 years . Both efforts were to prioritize the subset of odorants monitored within the mixture to identify “high priority” odorants for further analysis. For the sewer study, the upper bound concentrations were compared to the ODTC50 values from Nagata and Takeuchi . The ranking by OAV reduced the number of compounds from 31 to 8 “high-priority” odorants. Hydrogen sulfide and methyl mercaptan dominated across all sites , and other volatile sulfur compounds also ranked high for most sites. Diethyl sulfide, limonene, toluene and m,p-xylenes each ranked high for fewer sites. The limitations of such ranking are substantial. The choice of ODTC50 value is paramount and typically ranges over several orders of magnitude across studies, complicating OAV calculations and their utility as a ranking scheme. Although Sivret et al. acknowledged the enormous ranges of ODTC50 values available in the compilation by van Gemert in several graphs, the final ranking was performed using the single values found in Nagata and Takeuchi for simplicity. Presenting the final OAVs as ranges would have muddied the results. Finally, those compounds without ODTC50 values available in the literature were dropped from the ranking, essentially rewarding lack of data for potentially odorous compounds. As with any study of mixtures, the analyzed odorants are only a subset and may not reflect the overall, total odor experienced by residents. Risk assessment is often predictive rather than retrospective. The impacts of a proposed oil well in Hermosa Beach, California, were forecast . Except for upset conditions, the anticipated negative health outcomes were largely nuisance-related . The evaluation concluded that the oil well would have no substantial effect on community health, which was based in part on a risk assessment of hydrogen sulfide as the odorant of primary concern. The results of the modeling indicated that fugitive emissions from normal operations could produce concentrations greater than the odor threshold without mitigation, which would reach nearby residences. Concentrations could be as high as 6 times the odor threshold, primarily driven by hydrogen sulfide. The acute REL for hydrogen sulfide would only be exceeded, according to the model, during accidental or unplanned release. Odor impacts from normal operations were, therefore, considered potentially significant without mitigation, so an Odor Minimization Plan was required.Over years, the residents of communities north of Denver have complained of intermittent, unpredictable “tar” and “asphalt” odors. The symptoms in Globeville, Colorado, included burning eyes and throat, headaches, skin irritation and sleep problems. A USEPA funded environmental justice study was conducted there in 2012 . Air samples were collected from locations near potential sources and within the community over a period of 7 months. Out of a list of 23 analytes, the most prevalent were hexane, heptane, benzene, toluene, m,p-xylenes and naphthalene. The maximum concentrations were below odor and toxicity thresholds , except for the carcinogens , which were considered within normal ranges for urban air.