Interpretation of and extraction of useful information from irregular fluctuating industrial microbial counts

Irregular fluctuating patterns of industrial microbial records arise becaus e certain factors that affect them vary in a random fashion. If the counts are independent, and their series stationary with no trend or periodicity, a probabilistic model based on an asymmetric distribution function can be u sed to estimate the frequency of counts exceeding any given magnitude. Indu strial records of standard plate counts (SPC), anaerobes, and thermophiles in 4 dairy products were characterized with the lognormal distribution. Man y of the records had weak but significant autocorrelations for small lags, indicating that successive lots had similar microbial loads. Nevertheless, in almost all cases the method provided reasonable estimates of future coun ts frequencies, which were confirmed with fresh data.