[en] This work aims to formulating heat transfer processes via stochastic models. The bartlett formalism is introduced for treating the processes under analysis. The basic mathematical algorithms connected with the numerical description of stochastic variables are defined: factorial moments and cumulants, auto- and cross-correlation functions, auto- and cross-spectral density functions. Every thermal process is analyzed through the stochastic formalism previously introduced and the physical meaning of the analytical solutions found for each process is examined. The relevant features of the solution are collected in a synoptical box and theoretical formulae are given for the interpretation of heat transfer measurements in nuclear reactors