In this paper we analyze the additive hazard distribution modelling in the network system with non-constant nodes’ immunities. Nodes’ immunities depend on number of cycle and unknown parameter that is defined as random variable. Bayesian approach is applied for the updating of the estimate of this random variable. In this study we are interested in how many cycles can system work under influence of the hazard in the network or how many cycles are required to reduce amount of the nodes’ hazard to the safe level. Obtained results can be used in the prediction of system’ lifetime and accident analysis. Numerical experiment was performed to illustrate application of developed algorithm.