Predicting heat process efficiency in thermal processes when bacterial inactivation is not log-linear

The food industry widely uses the F-value which considers microbial log-linear inactivation, while microbial heat inactivation may result in a non-log-linear inactivation pattern due to genetic or phenotypical heterogeneity. This may yield discrepancies in predicting microbial heat inactivation under dynamic conditions of heat treatment. In this paper, we suggest the calculation of the equivalent time of heat treatment at a given temperature to overcome these constraints. To validate our proposal, the heat inactivation of Bacillus pumilus, showing non-log-linear behavior, was predicted for 4 different heat inactivation profiles and bacterial enumeration was performed to determine whether prediction errors were acceptable. When the proportion of residuals in an acceptable zone from 1 log (fail safe) to 0.5 log (fail dangerous) was greater or equal to 70%, the model was considered as acceptable for predictions of the tested data. The new approach gave four different temperature profiles, with 96, 85, 85 and 100% of the residuals in the acceptable zone, indicating satisfactory prediction. Thus the proposed practical alternative to simulate microbial heat inactivation kinetics is able to extend the F-value to non-log-linear inactivation patterns.