Complexity of technical systems, such as NPPs, implicates the necessity to undertake efforts for assessment of reliability of equipment, especially critical equipment influencing the reactor operation safety. Therefore, tasks associated with investigation of regularities of variations of parameters of equipment and the processes of their approaching to the conditions of equipment failures, and with the development of methodologies and algorithms for obtaining quantitative reliability indicator values with respect to progressive (parametric) failures become important. Such task is addressed in the present paper as applicable to ShAFDR-32M coolant flow rate sensors of RBMK-1000 reactor. Analysis of statistical data obtained during planned diagnostic measurements of two determining parameters of functionality of ShAFDR-32M coolant flow rate sensors (minimum values of negative half wave of the amplitude and standard deviation for the rotation period of the flow rate sensor ball) allowed developing the mathematical model of sensor parametric reliability. The process which is the superposition of elementary regeneration process and the stochastic process with independent increments will be understood as the mathematical model of coolant flow rate meter. Investigation of mathematical model of coolant flow rate sensor functioning reliability allowed obtaining in closed form the correlations between the average time of the flow rate meter operation before crossing of the preset boundary by each of the determining parameters and the probability of failure-free operation of the flow rate meter in asymptotic setting without introducing any assumptions with regard to the laws of distribution of random values.The results obtained can be easily generalized to embrace the case when the dimension of the vector of determining parameters is larger than two. Results of the investigation are applied in the calculations of quantitative indicators of parametric reliability of flow rate sensors.