Stored grain is susceptible to infestation by stored-product insects. Synthetic chemical insecticides and fumigants are being restricted due to health and environmental concerns, and insect resistance to the chemicals is developing. Because insects die when exposed to elevated levels of carbon dioxide (CO₂), modified atmospheres may be a viable alternative to chemical control methods.

The visible bin openings of two full-size welded-steel hopper bids were modified to improve the gas-tightness of the bins. Mean CO₂ concentrations improved from approximately 10% before the bins were sealed to almost 50% following the final sealing technique. These improvements are significant because the amount of CO₂ used was constant for all experiments. The sealed bin retained approximately 79% of the CO₂ that was initially added.

For fumigation in a grain-filled bin, the addition of a large volume of gaseous CO₂ had to be offset by a release of air from the bin. Purging was found to be most practical if dry ice was allowed to sublimate inside a sealed box outside the bin and the gaseous CO₂ was ducted into the head space of the bin. Air was released through a purge valve at the bottom of the bin. For five experiments, purging efficiencies ranged from 69 to 92%.

Retention efficiencies ranged from 55 to 82% during 10-d fumigations, but only from 28 to 42% during 4-d fumigations. The extra dry ice added during the 4-d fumigations did not produce the desired increase in CO₂ concentration, and consequently, retention efficiencies declined. Mortality of caged adult rusty grain beetles, Cryptolestes ferrugineus (Stephens), was 100, 99.8, and 99.7% in three fumigations of 10-d duration and 95.3 and 79.8% in two fumigations of 4-d duration. Fumigations of 10-d duration should be promoted over fumigations of 4-d duration if only an initial application of dry ice is to be used. Based on the published mortality data for C. ferrugineus exposed to elevated levels of CO₂, an equation was found that predicts the required exposure for any CO₂ concentration at a temperature of 25°C. Because CO₂ concentrations decayed during fumigations, a procedure was developed to apply the equation cumulatively on short intervals. The lethal exposure time was calculated based on the CO₂ concentration observed during each interval. A ratio of the interval to the lethal exposure time was calculated and summed over al intervals to give the cumulative lethality index. When the cumulative lethality index equals 1.0, complete insect mortality should occur. Calculated cumulative lethality indexes compared well with the observed insect mortalities in this research.

Gas-tightness varies from bin to bin. Gas loss rate was related to the pressure decay time through a common factor of leakage area to allow a CO₂ concentration profile to be generated for any bin prior to the start of a fumigation. With knowledge of the CO₂ concentration profile, the length of time required to achieve complete mortality of C. ferrugineus can be calculated. Predicted rates of CO₂ loss compared well with observed rates of loss.