IntroductionBulk railcars are a common method of moving commodities in the USA. Allowances are given for the practice of treating railcars with fumigates during transit because the routes are limited access and not on public roads. Recent technology has become available for monitoring phosphine gas (PH3) fumigation on railcars which logs the phosphine concentration and temperature of the test point in the railcars.Materials and MethodsTwo hopper bottom railcar shipments of corn grit were monitored for phosphine during 8-day transit from mill to processor. Several phosphine-sensing units were used in each railcar and spaced across the top layer. Mathematical modeling of the railcar fumigation was carried out using computational fluid dynamic software. Because access to lower depths in the railcar was not available, supplement experiments were performed with small columns of corn grits (2.5 m height x 0.55 m diameter) to test for phosphine at greater depths. Also, in the grain columns, bioassays of both phosphine susceptible and resistant, adult Rhyzopertha dominica (F.), lesser grain borer, and Tribolium castaneum (Herbst), red flour beetle, were included at the 0 cm, 25 cm, and 60 cm below the surface.ResultsThe phosphine concentrations in the railcar headspace varied with time with phosphine spiking over 1600 ppm and gradually settling to over 300 ppm at the end of the 8 days. Total gas dosage was estimated as concentration*time (CT) over the 8 days as 115,000 and 125,000 ppm*h at the top of each railcar. The supplement grain column fumigation tests found significant phosphine penetration into the column at 2 m depth with ~380 ppm after 2 days which reduced to ~260 ppm after 8 days, and all insects, at all locations, were dead after 8 days. The CFD simulation models were shown to provide estimates of the phosphine concentration and distribution which matched well with the observed data, validating the CFD approach as a useful tool.DiscussionThe simulation models were shown to provide estimates of the phosphine concentration and distribution which matched well the observed data, validating the CFD approach as an efficient tool for future planning and analysis of similar fumigations.