Comparison of helium leak test and vacuum leak test using canned foods: collaborative study.

Two can leak tests were compared by 7 collaborators. In the helium leak test, pressurized helium is applied to the outside of the container, and a headspace gas sample from the can is then analyzed for the presence of helium. The vacuum test is described in the Bacteriological Analytical Manual. Ninety No. 303 cans of creamed-style corn, green beans, carrots, fruit cocktail, and whole-kernel corn were shipped in 3 groups. Two groups of 30 cans had 10 dented flat cans, 5 flat controls (nondented), 10 dented swollen cans, and 5 swollen control cans (nondented). The third group had 10 dented swollen cans and 5 swollen control cans. Of 600 cans analyzed, 37 (6.2%) were deleted from the analysis because results were not available for both tests. One laboratory was constrained by scheduling to analyze 15 of 45 swollen cans. The helium leak test found 12 (13%) positives of 92 nondented swollen cans. One pressurization test yielded 7 of those 12 positives. Of the 400 dented cans sent as possible leakers, the helium test found 267 positives, and the vacuum test found 181. Five of the 7 analysts had significantly (alpha = 0.05) higher percent positive helium results. One analyst found more leakers by the vacuum leak test. Both tests found fewer positives in the swollen dented cans than in the flat dented cans. After exposure to pressurized helium, all cans with greater than 8 psi headspace pressure were positive helium leakers. The method was adopted official first action.