The physical environment also plays a role. Varying wind direction and speed lead to the transitory nature of odors, and multiple sources in the vicinity lead to difficulty in source attribution. Even temperature and humidity play roles in the perception of odor, which is often overlooked during exposure sampling and analysis. In addition to the large number of chemical compounds present in malodorous air, their typically low concentrations challenge the limits of even the best instruments . Known as the “odor gap,” the human nose can usually detect odors well below analytical instrument detectors’ capabilities . Methods that use human panels to evaluate odors have been standardized over the years and can work well in parallel with traditional analytical instrument methods. The vision is to have analytical instruments that completely mimic the human nose and sense of smell.The measurement and evaluation of exposure to conventional air pollutants is considered more evolved than that for odors . The framework and methodology applied to conventional air pollutants – risk assessment – offers grounding principles and useful conventions that have evolved over time. Both fields evaluate human responses to chemicals in the air. Although risk assessments are often predictive of future events, they may also be conducted retrospectively as an investigative technique.Risk is, by definition, is the probability of an adverse outcome and its severity. For chemical exposures, risk is a function of hazard and exposure . The fundamental framework for risk assessment was established in the 1980s . Figure 3.1 provides an overview of the various steps. These steps begin with the generation of basic information, proceed through identifying the hazards of the chemical under evaluation, predicting how adverse effects vary with dose,greenhouse tables and end with combining that information with exposure data to determine the incidence of adverse effects in a population.
Beyond risk assessment, and beyond the scope of this paper, is subsequent regulatory, management and communication steps based on the risk assessment’s output and other factors. Given the variety of information required in a risk assessment, the field is truly multidisciplinary. The data and assumptions made along the way are evaluated for how much uncertainty they contribute to the results. Often an order of magnitude or more of uncertainty and variability are inherent in the output, which needs to be explained transparently to not “over sell” the results with a false sense of precision and accuracy.A pragmatic approach to risk assessment is to first conduct a screening-level assessment based on crude approaches likely to overestimate risk. If the risk is found to be reasonable from such an approach, no further work is necessary. If not, then a more detailed, refined assessment is conducted. For the exposure assessment , the focus of this paper, a conceptual model guides the evaluation. The conceptual model traces the origin of the chemical , indicates how it is released, allows for transport of the chemical, includes possible routes of exposure, and indicates who might be exposed . Odors are released from a variety of sources, travel through the air and then are inhaled by local populations. Risk varies across a population due to biological differences , culture, lifestyle, level of exposure and prior exposures. To protect vulnerable sub-populations, a safety factor is usually applied. Perhaps the greatest challenge for both odor assessment and risk assessment is mixtures. We are exposed to a wide variety of chemicals through food, medicine and the environment, yet risk assessment often focuses on a single chemical in isolation. Odor assessment follows suit, focusing often on only one odorant. Such an unrealistic approach is destined to produce highly skewed or biased results, probably in unknown directions . Odor assessment has the advantage of tests being performed by human panels, which can evaluate the whole mixture of the sample. Risk assessment relies on epidemiological reconstructions for human data.Risk assessment, however, has developed approaches for mixtures. A simple, screening level approach is to determine the risk-driver for the mixture. Adding up the individual effects is another crude approach. A simplifying aspect for odor exposure assessment is that human olfaction has evolved to differentiate between only a few significant stimuli.
Typically, around 3 or 4 odors are sensed at a time, which decreases the complexity of the mixture . Those odorants that trigger intense, familiar or unpleasant sensations are more likely to be noticed while the remainder are lost in the signal “noise” or sensory filters. Or this limitation may due to inability to name a substance, rather than failure to detect the difference between odors . Both risk assessment and the evaluation of odors suffer from high degrees of uncertainty and variability. The personal nature of odor perception introduces fundamental variability. The health effects evaluated in risk assessment have a similar range of variability due to the biological variability of humans, which is increased further by the extrapolation of animal studies to humans. Therefore, each health effect benchmark value, such as a toxic reference dose, is typically presented with one significant figure due to the inherent uncertainty, which typically spans an order of magnitude. Exposure results, too, are uncertain due to modeling assumptions or analytical imprecision, as well as sample collection issues. In reality, one significant figure is a misrepresentation, and a range would be more accurate. Making judgements using ranges, however, is difficult so single values are typically used. A sensitivity analysis helps show the possible range of results.Transparency each step of the way is paramount, otherwise overconfidence in shaky results may occur. Both the best practices and draft guidance include a tiered approach to odor evaluations. Such has long been used in risk assessment to streamline the work. First, a screening-level evaluation is performed using crude assumptions and approaches. If the exposure is deemed acceptably low, there is no need for further work. The same applies to odor investigations. If a straightforward evaluation by an air inspector identifies the source and resolves the issue, no complex further investigation need ensue. In both cases, if the screening-level approach identifies concerns, then a detailed analysis is undertaken.Describing an odor in detail is often difficult, so most complainants start with saying “something smells bad” and then struggle to give further details. Unlike other senses with broad vocabularies, smell is anchored in the source of the odor and the person’s history with that source. In a way, our sense of smell is learned.
Attributing words and meanings to odors occurs over a lifetime and even changes over time. The food and beverage industry has attempted to make a science out of sensory description. Beer, wine and coffee are prime examples. Perfume formulation takes this to another level. To avoid complaints,vertical farming the drinking water industry has developed taste-and-odor assessment protocols.Environmental odors are typically mixtures of chemicals . The rare exception is the release of a single odorant from a chemical industry facility. The various odorants within a mixture trigger the olfactory sense in “concert” similar to the various notes in an orchestral piece of music. The perfume and fragrance industries are built largely upon this principle. The interplay of odorants in a mixture can be complex, with both synergistic and antagonistic effects taking place. Perfume has the function of covering up other odors. In odor terminology, this is called “masking.” Landfill and bio-waste sites are known to use scents such as “cherry” at their perimeter , yet in an evaluation of commercially available masking products only 4 out of 26 were able to mask odors successfully . All 4 were neutralizing agents that reacted with odorants. Within an environmental odor sample, certain odorants may mask others. Only upon dilution to a point where the major odorants are no longer perceptible are the minor odorants noticed. This dilution effect has been termed “peeling the onion” , where one layer of odor leads to another. Further discussion of this effect is in the section on odor intensity. The odorants within a mixture are subject to the same physicochemical processes and dispersion as any conventional air pollutant. The same exposure models, such as fate and transport, apply; however, the identities and concentrations of the individual odorants are often unknown, rendering such modeling impossible. To get around this issue, a pseudo-concentration approach has been developed, which is discussed in Section 4.2. The overwhelming majority of the molecules in air are odorless. These include nitrogen, oxygen, water, hydrogen, helium and carbon monoxide. Rather uniquely, carbon dioxide is odorless until it reaches 200-fold above background levels , at which point is triggers the nasal trigeminal receptors rather than the olfactory receptors.Colors have agreed-upon descriptions, and graphic artists often use Pantone® numbers as specific identifiers. Musical notes have frequencies assigned to them and arranged into scales . Odorants, too, have descriptors, known as “notes,” the term used in ISO 5492:2008 . For example, “fishy,” “swampy,” “rotten egg,” “pungent,” or “tingly” are odor notes. An atlas of panel-derived odor notes has been published . The odor note, however, may change with the concentration . Hydrogen sulfide at levels above 20 ppm changes from its characteristic “rotten egg” odor note to a “sweet” odor note, and at even higher concentrations, which are toxic, hydrogen sulfide becomes odorless. The response to an odor is highly personal and depends on “odor memory” – previous exposure and knowledge about the odor source . Common descriptors associated with specific odorants, however, may aid in determining the source of an odor.
Odor wheels have been developed for specific odor notes associated with certain sources, such as landfills, composting and WWTPs . Odors as mixtures make assigning odor notes more complex. As with wine tasting, several dominant notes may be present, along with several subtle notes. These, too, change as the mixture is diluted or ages, or as temperature and humidity change.As with sound and color, some odor notes may be perceived as pleasant or unpleasant. This is the “odor hedonic tone,” also known as the acceptability of the odor. Dravnieks published on this topic, and a scoring system is named after him. Odor hedonic tone is a highly subjective determination, open to large variation across a population and appears to be learned rather innate . Odor hedonic tone varies as the odorants increase or decrease, sometimes progressing through flip-flops between pleasant and unpleasant .Odor intensity – the magnitude or strength of an odor – has received considerable attention. Unlike odor notes and hedonic tones, which can be fairly subjective for the untrained, odor intensity is pursued as a quantifiable, even scalable, attribute of odor perception. The belief is that odor intensity is akin to brightness or loudness, which are quantifiable through physics, yet odors are a chemical sense with accompanying complexities. Nonetheless, two approaches have been attempted: assigning words or numeric scores to intensity levels, or determining the amount of dilution required until the intensity is no longer detectable. For a single odorant, intensity appears to be linked to the odorant’s concentration. In mixtures, such a link is tenuous or absent. Although odors are typically mixtures, it is much easier to study individual odorants. The mathematical connection between odorant concentration and perceived intensity is called the Steven’s power law . Whether the concentration is used directly or divided by a reference concentration does not impact the relationship between intensity and concentration. The inclusion of an intercept, however, fundamentally changes the slopes of the lines and does fundamentally alter the relationship. The original Weber-Fechner law included no intercept . Researchers added the 0.5 intercept to account for “the definition of the odour threshold concentration which states that 50% of the panellists perceive weak odour while the others perceive no odour” and proceeded to use an intensity scale that ranged from 0 to 5 . Other researchers also used the equation with the 0.5 intercept, but the intensity scale ranged from 0 to 6 . The effect of the 0.5 intercept in both of these studies was to assign an intensity score of 0.5 to the ODTC50 concentration, which had nothing to do with the percent of the panelists perceiving or not perceiving and odor. Another researcher allowed the intercept to float uniquely for each odorant and used a 0 to 12 intensity scale , which scale been used in flavor and drinking water profiling .