Effects of weather conditions on sulfuryl fluoride and methyl bromide leakage during structural fumigation in a flour mill

A validated Computational Fluid Dynamics (CFD) model of structural fumigation was used to perform 54 fumigation simulations (half with sulfuryl fluoride, SF, and half with methyl bromide, MB) using recorded hourly average weather data for 4 July and 1 September at Indianapolis International Airport for the years 1996–2006. These weather data were used as the simulation boundary conditions. Although the fumigations were simulated for the same time periods, the results indicated substantial variations in the fumigant leakage rates. The 11-year averages of wind speeds on 4 July and 1 September were 3.8 and 2.4 m/s, respectively, resulting in half-loss time (HLT) differences between the fumigations on the two days. On average, the HLT was 4 h lower on 4 July than on 1 September. When comparing the results between individual fumigations, substantial differences in the fumigant leakage were observed, regardless of the fumigant type and whether or not the fumigations were performed at the same time period. This implied that using past fumigation data as the primary means for evaluating the structural sealing quality of a current fumigation is not adequate. Predictions of fumigant leakage rate and fumigation performance should incorporate quantifiable sealing effectiveness and weather information for the planned fumigation period. Comparisons between SF and MB fumigations indicated that under exactly the same weather conditions and fumigation practices the leakage characteristics of SF and MB do not differ. In practical situations where the dosage requirements for SF and MB are typically not the same, however, the leakage rates of SF and MB fumigations could be different due to the buoyancy effect. Nevertheless, the magnitude of the difference may or may not be significant depending on other factors such as sealing quality, wind speed and direction, and ambient temperature. The effects of these factors should be further quantified.