CUSTOM SOFTWARE FOR LOCATING PROBABLE SOURCES OF ODOR USING COMPLAINT AND METEOROLOGICAL DATA

When odor complaints occur in a community questions often arise about the source of the odor, and the validity of the complaints. Plant operators may be defensive and dismiss complaints, or attribute them to other potential odor sources in the area. Odor issues center on complaints. Complaints both define the problem and are a measure of it – there is no odor nuisance if no-one complains, and the more people complain, the greater the problem is. Wastewater treatment plant operators commonly track the number of complaints per year as a measure of their progress in addressing an odor nuisance concern. Complaints can also be viewed as valuable data which can help to identify the sources or causes of an odor nuisance.The analysis of complaint data is usually very rudimentary. Typically, reference is made to wind data to check the validity of complaints based on wind direction. Dates and times of complaints may be compared to plant operations data to see if complaints correspond to particular operating conditions. Locations of complaints are plotted on maps to hopefully identify patterns that would lead to some insight into the odor nuisance. Seasonal patterns in complaints may be discerned from monthly or weekly tabulations. While these simple approaches have some value, more insight into the odor nuisance concern can be obtained from a more sophisticated analysis.This paper describes a program based on a procedure described in the early 1990s by Tapper and Sudbury, which uses odor complaint data to identify the likely source or sources of odor. The program may be thought of as dispersion modelling in reverse. Rather than predicting the impact of a source on a community, this program takes the impact data, i.e. complaints, and calculates probable locations for the source. The steps involved include first defining the study area, and dividing this area into a grid. A database of complaint information, including location (x, y coordinates), date and time is constructed. This is matched with a database of meteorological conditions, including wind direction, wind speed and atmospheric stability. For every complaint record, and for every grid point in the study area, a simplified Gaussian model is used to determine whether an odor source at that location could account for that particular complaint. At the end of this procedure a graphical representation (contour plot) is shown on a map of the area indicating the locations which would account for a given percentage of the complaints.This program was used in a study at a wastewater treatment plant in the summer of 2002 to analyze historical and new complaint data as well as data obtained from a community odor survey. Output clearly indicated the wastewater treatment plant as the primary odor source in the area. There were indications of secondary odor sources in the area, possibly associated with the sewage collection system.