A Simple Concept Allowing the Prediction of Microbial Inactivation Under Non-Isothermal Process, Taking into Account Non-Log-Linear Inactivation Kinetics

Introduction:  During pasteurization and sterilization, microorganisms are exposed to non-isothermal process, as a consequence of the heating penetration. Nevertheless, nowadays, microorganism heat parameters are obtained under isothermal conditions; and furthermore, the prediction becomes more complicated, when the inactivation kinetics are non-log-linear. Thus, the investigation of thermal models that may accurately predict heat inactivation for non-isothermal heat treatments is a topic of interest for food industry.

Purpose:  In this study, we investigated the performance of two published models (Valdramidis et al., 2011 and Peleg et al., 2000) and one new one.

Methods:  Briefly, the proposed model converts the non-isothermal profile into an isothermal profile for a given temperature and could be combined with both a linear or Weibull primary model for further prediction. To compare the different models, heat resistance parameters of Bacillus pumilus E71 were estimated from 11 non-log-linear inactivation kinetics (68°C-101°C), obtained under isothermal conditions. Thereafter, inactivation kinetics were acquired under seven temperature profiles and performances of the simulation were assessed statistically (RMSE, A_f and B_f).

Results:  Calculated RMSE varied from 0.14 to 2.25 for all predictions and for six temperature profiles. The new model combined with a Weibull model gave the lowest forecasting error. According to the bias factor, Peleg model ‘fail-safe’ for six temperature profiles, Valdramidis ‘fail-dangerous’ for all, and the proposed model over and under predicted three and four times, respectively. The accuracy factors closest to one, obtained for the new model, were mixed; indicating that the average estimate was more accurate for this model. All these results underlined that the proposed model was more robust than the others for prediction of heat inactivation under non-isothermal treatment.

Significance:

In conclusion, the present study provides a new model, which could be used to predict the microbial heat inactivation under non-isothermal process and, thus, can lead to effective management systems for the optimization of the pasteurization or sterilization steps.