Industrial odour control
4 Measurement of odour
| 4.1 | Methods |
| 4.2 | Measurement procedure |
| 4.3 | Sampling odours |
4.1 Methods
The threshold value of a sample is determined by means of an odour panel, determining the dilution where 50% of the observers have a conscious sensation of odour. The sample is said to be diluted to the odour threshold value. Most odour panels consist of 6-10 panellists.
Determination of odour thresholds
In Denmark two methods for determination of odour thresholds have been used in particular (MØLHAVE, 1980). One is a high volume method developed in Sweden (LINDVALL" 1970), the other is a method developed in the USA on the basis of the triangle olfactometer (DRAVNIEKs, 1975). Both methods were used to determine dilution factors for industrial exhaust air. However, in general odour measurements with different equipment and different procedures do not yield the same results, because, as explained above, the sensation of smell and thus, the odour threshold value depends on a number of conditions, i.e. air speed, temperature and humidity, and the background odour of the equipment/room. Another important factor is the degree of dilution before air reaches the odour sensitive regions of the nasal cavity. If the air speed in the apparatus is low, the degree of dilution will vary during sniffing.
Determination of odour threshold value
The determination of the odour threshold value of a substance depends on the measurement method used. Therefore we cannot in general compare threshold values found in literature with measurements we have made ourselves, unless the methods used were identical.
Measurement methods should fulfil the following general requirements:
- Air should be dosed dynamically.
- The flow rate of sample gas to panellists should be 20 l/min.
- The dilution steps should not exceed a factor of 2.
- The odour presentation should be short, to avoid adaptation.
- Panellists’ answers should be by the forced choice principle or other principles to avoid anticipation biases.
- Loss of odorant and contamination should be eliminated by correct choice of material etc.
The Danish EPA is aware that these requirements can hardly be fulfilled by all currently used measurement methods.
Odorant unit
The results of threshold determinations can be expressed by a figure describing the odorant unit, LE, defined as the amount of odorants mixtures which distributed in 1 cubic metre of air results in odour intensities corresponding to the threshold value determined by a given procedure. The odorant unit is, thus, defined by a physiologically measured amount of substance.
Odorant concentration
If as a result of the threshold value determination, a sample must be diluted 100 times to reach the threshold value, the odorant concentration is 100 LE/m3. Below the symbol C is the odorant concentration.
4.2 Measurement procedure
Below a measurement procedure is described which must be used to determine the odorant emission in exhaust air.
Reference substances
Odour thresholds are determined using l-Butanol and hydrogen sulphide as reference substances. The panellists must regularly determine the threshold concentration for l-Butanol and hydrogen sulphide. Threshold value concentrations of 0.05 ppm for l-Butanol and 0.0006 ppm for hydrogen sulphide were used as a basis for assessing the panellists' sensitivity.
Sensitivity factors
If a panel determines the odour threshold concentration at Cb (ppm) and Cs (ppm) respectively, the sensitivity factors relating to l-Butanol and hydrogen sulphide are said to be
and
respectively, and the resulting sensitivity factor is defined as
If the odorant concentration C is determined by this panel, the odorant concentration C5o is used to calculate the effective outlet height and/or dimensioning of cleaning measures. C50 is calculated by
C50 = C/P
So with a dilution of C50, the odorant concentration of a sample with concentration = C50 corresponds to the odour threshold value determined by a panel with a sensitivity factor P = l.
Selection of panellists
The panellists should be selected along the lines described by MØLHAVE (1980). The following general requirements should also be applied:
The panel must consist of at least 6 persons between 18- 50 years.
- All age groups, between 18 and 50 years should be represented.
- The panel must be able to determine threshold values of l-Butanol and hydrogen sulphide with reproducibility ± 30%.
- The geometric dispersion of panellists' responses should be less than 1.5 for l-Butanol. The geometric dispersion is determined by means of measurement values around the median. At least 2 measurement points different from 0% to 100% should be used.
4.3 Sampling odours
Odour sampling, collection and storing
Samples for determination or the odorant emissions must be taken according to the directions described in Guideline No. 7/1974 (Miljøstyrelsen, 1974) on measurement of gas emissions. Collection and storage of samples are described by several authors (i.e. MØLHAVE, 1980). Change or contamination of the sample during collection or storage can for instance be avoided by using polyvinyl fluoride plastic bags (PVF-TEDLAR) connected to the measurement point by a teflon pipe not more than 2 m long. The bag is placed in a rigid container, which is evacuated, allowing the tedlar bag to inflate. If exhaust air is hot or contains dust or a liquid, it must be cooled, filtered or dried without loss of odorant. It may be difficult, and requires a thorough physical/chemical evaluation of the conditions. The durability or the air sample should be examined by physical/chemical measurement methods, allowing the maximum allowable storage period to be fixed (normally not over 24 hours). The sampling time should be 1-5 minutes with suction speed = 4-20 l/min. The samples should be taken when the odorant emission is at a maximum in normal conditions of operation. The volume of the odorant emission is calculated as the arithmetic mean of measurements made with maximum odorant emission. Further details can be obtained from the Danish EPA or consultants working with odour measurements.