Is an “odour unit” a scientific measure sufficiently reliable to be a legally enforceable condition in an environmental approval? The best analysis to date is a decision of the British Columbia Environmental Appeal Board, West Coast Reduction Ltd. v. Greater Vancouver (Regional District). According to the Board, odour units are not scientifically reliable enough for this purpose.
West Coast Reduction operates a rendering plant, which provides an essential public service, but inevitably has some odours. A community campaign to “stop the stink” created an “explosion” of complaints, which persuaded the Greater Vancouver Regional District to impose stringent new requirements on the company’s air permit. Some of those requirements imposed odour unit limits on various parts of the plant. The company appealed, and called expert evidence about the subjectivity and unreliability of odour unit measurements.
The Board decided:
“… the Panel finds that the use of odour units in this context is not reasonable and appropriate. The notion that odour units can be used as an indicator of an environmental “smell” is simply too flawed to be used as a method of determining compliance, and is therefore not suitable for determining whether the environment is adequately protected.
332 To begin with, an odour unit is a dilution ratio. The mathematical definition of “ratio” is dimensionless. Therefore, to give an odour unit a “unit of measure”, is already predisposing it to a “mass”, which it is not, and is therefore arbitrary.
333 Further, the dilution ratio is equal to the volume of clean air divided by the volume of diluted air (or the diluted odour). In order to attribute a “measure” of odour units to a sample of air, human panelists are used. The Panel appreciates that this is considered the best, and possibly the only, means of measuring smell. However, the basis upon which the panellists are chosen, the use of n-butanol, is not without its problems, particularly when it comes to the correlation between sensitivity to n-butanol and to environmental odours.
334 The European Standard is based on an assumption that the performance characteristics as determined on reference materials are transferable to other odourants. Specifically, that there is a linear relationship between a person’s sensitivity to n-butanol and to other odours. If the person can detect between 20-80 parts per billion, they qualify to be a panellist. The response obtained to 40 parts per billion in n-butanol is the standard upon which other odourants are referenced.
335 The Panel finds that there is no credible support for this assumption in the context of the environmental odours at issue in this case….
that bias and subjectivity are present at many stages during the capture and analysis
process:
1) In order to assess accuracy of analysis one needs to compare the result with a standard, a true value. In olfactometry,
human panellists are used in place of analytical instruments. The difficulty in assessing the accuracy of the
odour unit is that there is no clear measure as to the true odour unit value because the response to the sensor (i.e., the
nose) varies among humans. Thus, bias and subjectivity is introduced at this point.
2) Odour units may be derived from the laboratory and from the field. Many times the values obtained from these
two different methods are used interchangeably, which is erroneous. Bias is again introduced at this point.
3) The human panel process does not account for the possibility of residual odour in a panellist’s nose after completing
multiple rounds with several presentations. Bias can be introduced at this point.
4) The Tedlar[TM] bags used for the samples emit odour powder plus solvent. The solvent is emitted from the bag
and could interfere with the measurement as well. Background odour from a bag is a problem if measured levels are
so low which causes interference with lab analyses. Thus, another bias is introduced….
345 Given that there are many steps in the process of attempting to calculate odour units which are problematic, and
which contain so many points of bias and subjectivity, the Panel finds that the ultimate number or value coming out of an
odour unit measurement cannot be relied upon as meaningful, particularly for the purposes of evaluating compliance with
a mandatory term of a permit.