[en] Current reliability assessments of safety critical software embedded in the digital systems in nuclear power plants are based on the rule-based qualtitative assessment methods. But practical needs require the quantitative features of software reliability for Probabilistic Safety Assessment (PSA) that is one of important methods being used in assessing the whole safety of nuclear power plant. This paper discusses a Bayesian Belief Nets(BBN) based quantification method that models current qualitative software assessment in formal way and produces quantitative results required for PSA. Commercial Off-The-Shelf(COTS) software dedication process was applied to the discussed BBN based method for evaluating the plausibility of the method in PSA