A hybrid FTA-Bayesian approach for leakage risk mitigation in molten salt storage tanks: Experimental validation and engineering solutions

• Developed a novel risk assessment model for Bolt-Gasket-Flange Connection (BGFC) leakage in molten salt storage tanks. • From a practical engineering perspective, consider the failure mechanism of leakage in the bolt-gasket-flange system caused by high-temperature conditions. • Incorporated Noisy-OR gates to enhance expert knowledge integration while reducing workload requirements. • Implemented fault tree analysis based on BGFC leakage failure mode classification. • Identified material defects, human factors, and operational conditions as primary leakage contributors through systematic analysis. The bolted gasket-flange connection (BGFC) is a vital pressure-retaining component in high-temperature molten salt storage systems for concentrated solar power (CSP) plants. This study proposes a novel hybrid method that integrates the analysis of failure mechanisms under high-temperature conditions, a fault tree with a Noisy-OR gate, and sensitivity analysis (SA) to evaluate BGFC failure risks. A fault tree was developed using the Noisy-OR gate method to calculate the minimal cut sets and top risk factors. Through Sensitivity Analysis (SA), it was determined that the key factors contributing to BGFC failure. Implementing this method is conducive to preventing BGFC failure and holds significant practical value for environmental protection, ensuring safe equipment operation, and preventing accidents.