Development of the OFFSET model for determination of odor-annoyance-free setback distances from animal production sites. II. Model development and evaluation

The OFFSET (Odor from Feedlots - Setback Estimation Tool) model was developed to estimate the setback distances from animal production sites in Minnesota. It is based on odor emissions taken from field measurements and an evaluated air dispersion model. The odor emissions of a site were estimated using odor emission rates that were the geometric means of odor emissions measured from 280 animal buildings and manure storage units on 85 farms in Minnesota. The odor-annoyance-free intensity level was set at 2 (faint odor) on a 0 (no odor) to 5 (very strong odor) intensity scale. An evaluated air dispersion model, INPUFF-2, was used to calculate setback distances from various animal farms for the set odor-annoyance-free level under six weather conditions that favor odor transport. Setback distances are presented in a graphic form as well as mathematically as a function of the total odor emission factor and the desired odor-annoyance-free frequency of the neighbors. Odor-annoyance-free frequencies between 91% and 99% are based on the average weather data for Minnesota from 1984 to 1992. Suggestions for odor-annoyance-free frequency selections are given. The OFFSET model also deals with residences located in different directions from a livestock site. Additionally, it can determine the odor occurrence frequency of a residence surrounded by several livestock sites. Comparing the setback distances obtained from the OFFSET model and the odor events reported by the resident observers, it was found that the OFFSET model does not overpredict odor transport distances under very stable weather conditions. By comparing the OFFSET predictions with the odor complainers' distances from swine farms, it was clear that their residences had high odor occurrence frequencies. The OFFSET model was also evaluated by comparing odor occurrences documented by the resident odor observers in the vicinity of eight livestock farms. It was found that although the model may describe the average neighborhood intensity correctly, a high variation in the observed odor intensities existed for all levels of predicted intensities calculated from the OFFSET. Further research is needed to improve the accuracy of OFFSET and also to improve the field odor measurement method by the resident observers to obtain reliable odor occurrence data. By comparing OFFSET with four other existing setback guidelines, it was found that the distances required by the other models fell in or below the 91% to 98% annoya