Abstract: Objective As a key power equipment in long-distance pipeline networks, the reliability of compressor units directly affects the safety and transmission capacity of gas transmission systems. In response to the problem that the failure mechanisms of various subsystems of compressor units differ and traditional methods are difficult to evaluate uniformly, this study conducts reliability research based on hierarchical modeling to achieve quantitative evaluation and dynamic characterization of reliability under multi-system coupling conditions. Methods A hierarchical reliability evaluation model is constructed, decomposing the compressor unit into system layer and subsystem layer. At the subsystem layer, reliability models for mechanical, thermal, and auxiliary systems are established respectively: the mechanical system is based on stress-strength interference theory, combined with finite element analysis and Monte Carlo methods, to calculate the reliability of key components such as impellers and bearings; the thermal system is based on polytropic compression theory and dimensionless performance curves, selecting surge margin, efficiency degradation, and abnormal states as key indicators, and using the analytic hierarchy process to determine weights to construct a comprehensive evaluation model; the auxiliary system is based on historical operation data of multiple units to establish an exponential distribution failure model for reliability assessment. At the system layer, structural logic relationships between subsystems are established, and by dynamically correlating subsystem reliability with key operating parameters, a dynamic calculation method for overall machine reliability suitable for variable operating conditions is proposed Results This research establishes a dynamic reliability calculation method for compressor units based on hierarchical modeling. For the subsystem layer, reliability models are established separately according to the characteristics of each subsystem; for the compressor unit system, a system dynamic reliability model based on the logical relationships between subsystems is established. Conclusion The proposed reliability evaluation and dynamic calculation method based on hierarchical modeling effectively characterizes the reliability evolution law of compressor units under multi-system coupling and variable operating conditions, which can significantly improve evaluation accuracy and dynamic level, and can provide theoretical basis and technical support for compressor unit operating status assessment, maintenance decision-making, and life prediction, having reference value for ensuring the safe operation of natural gas long-distance pipeline networks.