The occurrence of leaks fall into two categories, the first is leaks that randomly occur during a normal packaging run, and which are a measure of the overall process capability. These leaks cannot normally be attributed to any fault condition and occur at random intervals through the production run.
The other type of leaks are caused by a fault condition and are more recognized as a cluster of leaks or a stepped increase in the occurrence of leaking packs. That’s why manufacturers use leak detection testing methods, to identify these types of leaks.
Examples of random leaks would be products falling into the seal area, slight changes in the process conditions due to inconsistent running of the line, or a heavy crease forming in the packaging material and then disappearing again. Variation in film quality and general overall film quality will also produce random leaks through a run.
Clusters or stepped increases in leak occurrence can be a result of a failure in a heating element, failure of a thermocouple or a component failure, though these will need intervention to correct the problem. Contaminations build up on seal tools and web tension problems will create clusters of leaks that can appear and disappear during a production run.
Offline sampling is one such sampling regime, where the overall production capability can be measured. On an average line running at a throughput of 80 packs per minute, a well-managed sampling regime could sample 1 % of the product, which in an eight-hour shift could produce a sample size of 300 packs per shift. This will give a fairly good indication of the process capability, and be useful as a reference point from shift to shift performance.