Evaluation of measured and predicted odor concentrations around a meat packaging and rendering plant

ELBİR T., Dincer F., Muezzinoglu A.

ENVIRONMENTAL ENGINEERING SCIENCE, vol.24, no.3, pp.313-320, 2007 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 24 Issue: 3
  • Publication Date: 2007
  • Doi Number: 10.1089/ees.2005.0105
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.313-320
  • Keywords: odorous emissions in Izmir, odors from rendering, odor dispersion, ambient odor levels, CALPUFF, DISPERSION MODEL, EMISSIONS, DISTANCE, IMPACT
  • Dokuz Eylül University Affiliated: Yes


Odor levels around a meat packaging and rendering plant were predicted by using the CALPUFF dispersion model. The product of olfactometrically measured odor concentration and volumetric flow rate of the stack gas stream was used as emission strength of the source. Ambient odor levels were also measured at two sites downwind of the plant by olfactometry. This was possible, as odor concentrations were above 50 odor units per cubic meter (OU m(-3)), and this means that the concentrations were always exceeding the minimum 10 OU m-3 that can be safely measured by this technique. Odor samples were taken consecutively at two sites downwind of the source with 10 s duration each. These sampling sites were chosen so that they had continuous odor nuisance problems, which indicated that they were near the center-line of the plume. Between samples 10-s intervals were necessary to prepare the sampler for the next sampling bag connection. The overall sampling period was 10 min each at the two sites. Meteorological measurements were carried out throughout the odor sampling process at the sites. Wind direction and speeds were recorded at the end of each sampling and used in predictions made using the model. Predicted results were compared with the odor measurements carried out on the samples. Measured and predicted odor concentrations were tested to see the degree of correlation. The measured and predicted results at such very short averaging times correlated well, with a good agreement between the two series.