Applications of Probabilistic Engineering in Food Moisture Management to Meet Product Quality, Safety, and Shelf-Life Requirements

This chapter is devoted to the inclusion of the variability in a Monte Carlo-based probabilistic approach when using estimation models for moisture management purposes. The examples include deterministic and probabilistic estimations of: equilibrium water activity for multiple ingredients with different initial water activity; shelf-life of moisture-sensitive products; and estimation of the packaging moisture permeance to meet a required/desired product shelf-life. They also include estimation of the optimized package fill setting by risk analysis and estimation of the moisture permeance variability. The chapter presents a review of the information theory criteria used by food scientists and engineers to evaluate and select the most suitable mathematical model to be used for estimations. Its advantages are highlighted using as an example the selection of predictive isotherm and glass transition temperature models and their use to determine the critical moisture content for dried fruits.