Fumigation with carbon dioxide (CO2) can be a viable alternative to synthet
ic chemical control of stored-product insects, but storage structures must
be well-sealed for the method to be efficient. Because the degree of gas-ti
ghtness is not consistent for all sealed bins, it would be beneficial if th
e gas-tightness could be determined before the start of a CO2 fumigation. T
he hypotheses tested were that leakage area could be calculated from a meas
urement of pressure decay time and that gas loss during a fumigation could
be predicted from the calculated leakage area. Wheat-filled pilot bins with
holes of known areas were used to validate the relationship between pressu
re decay time and leakage area. The mathematical model predicted leakage ar
eas in the pilot bins with errors ranging from -17.5 to 23.1%. Predicted ga
s losses from pilot bins filled with polyethylene pellets were inconsistent
with the observed gas losses. Gas loss decreased as the initial CO2 concen
tration decreased.