Urban public facilities tunnels are vital underground infrastructures that transport electricity, communications, gas, water, and drainage systems. The safe and stable operation of these tunnels is directly related to the normal operation of the city and public safety. During long-term operation, these tunnels often face some common industry problems, such as the coupling of multiple risk factors, high uncertainty of accident occurrence, significant changes in dynamic risks, and lack of a large number of historical fault records. In order to solve these problems, this paper proposes a quantitative operation safety assessment method for urban public tunnels based on the fuzzy Bayesian network framework. By means of systematic literature review, field engineering investigation and semi-structured expert interviews, this method comprehensively identifies safety risk factors from four dimensions: structural integrity, equipment operation status, external environmental impact, and operation and maintenance management capabilities. A multi-level index system is constructed. On this basis, fault tree analysis is used to clarify the causal propagation logic and accident evolution path between risk factors, and the fault tree model is directly mapped to the Bayesian network topology. At the same time, combined with fuzzy set theory and expert experience, the expert evaluation in the form of language is transformed into trapezoidal fuzzy numbers. After defuzzification, the prior probability and conditional probability parameters of Bayesian network are obtained, which effectively solves the technical bottleneck problem of accurately obtaining parameters in the case of insufficient historical data. This method has been verified in a real urban public tunnel project. The overall safety risk probability is calculated by forward reasoning, and the key risk factors leading to the accident are located by reverse diagnosis technology. On this basis, targeted risk control measures are proposed. The research results show that this method can scientifically quantify the safety risk level of public tunnels, effectively solve the ambiguity and uncertainty problems in the evaluation process, and provide theoretical support and engineering reference for the safety management and risk control of urban public tunnels.