A study of the randomly fluctuating microbial counts in foods and water using the Expanded Fermi Solution as a model

Abstract

Randomly fluctuating industrial microbial count records, with and without zero counts, were simulated with a version of the Expanded Fermi Solution, originally developed for risk assessment. The basic assumption has been that each individual count is determined by the multiplicative effect of several random factors, which augment or suppress the microbial population size, and in the case of sporadic pathogens, determine the probability of their initial presence too. Records were generated by a series of Monte Carlo simulations in which the factors were specified by ranges and their values chosen randomly within them. The process has been automated and posted as a freely downloadable Wolfram Demonstration on the Internet. The program allows the user to enter and alter the series length, parameters' ranges, and count level deemed dangerous with sliders on the screen. The display includes the chosen factors' ranges, the corresponding generated count record and its histogram, and an estimate of the risk of surpassing the dangerous threshold. Where the record contains no zero counts, the histogram is accompanied by the lognormal distribution, which naturally emerges from the fluctuations' mathematical model. Once the factors are identified and their ranges specified, the method could be used as a tool to analyze, compare, and quantify microbial risks in foods and water. © 2011 Institute of Food Technologists ®.