CORESTA Meeting, Smoke Science/Product Technology, 2013, Seville, ST 47

LIP cigarettes: Proposal for an alternative sampling design

VERRON T.(1); CAHOURS X.(1); COLARD S.(2)
(1) SEITA, Imperial Tobacco Group, Fleury-les-Aubrais, France; (2) Imperial Tobacco Limited, Bristol, U.K.

Low ignition propensity cigarettes are intended to increase the chances of self-extinguishment under laboratory conditions following the protocol detailed in either ISO 12863:2010 test method or ASTM E2187-09. The safety requirements state that a cigarette batch is certified to have passed the test when 75% or greater of the cigarettes, coming from a sample size of 40 cigarettes, self-extinguish before burning to the tipping paper. However, the key point is to know the link between the percentage of compliant products obtained with the sample (40 cigarettes) and the genuine percentage in the entire production batch. In other words if a batch had a percentage of compliant cigarettes of 75% then, what would have been the probability with a sample of 40 cigarettes to observe less than 75% of self-extinguishment?

To answer this question, the statistical properties of the sampling design used to certify the cigarettes was investigated, and the sampling theory was approached to determine if the application of a statistically equivalent alternative sampling design was possible. From the understanding of the sampling theory consequences, an improved alternative sampling design can be proposed. This alternative has the benefit of reducing the mean testing time to certify the products without any loss on the statistical performance.