Onboard Carbon Capture System (OCCS) realizes carbon emission reduction of ships by capturing carbon dioxide (CO₂) exhaust gas generated during ship navigation and liquefying and storing it until it is offloaded ashore for professional treatment. Since the offloading of liquid carbon dioxide (LCO₂) has a certain risk, and the operation is complicated and requires high professional technology, this paper launches a research on the risk assessment method of liquid carbon dioxide offloading. Based on the storage conditions of LCO₂ in the shipboard carbon capture system and the unloading method and process, we analyze the content of the unloading risk assessment and the acceptable level. A fuzzy system is introduced to accurately reflect the degree and level of liquid carbon dioxide offloading risk and form a liquid carbon dioxide offloading risk assessment method. Collect the risk data in the actual unloading process of liquid carbon dioxide, and use the fuzzy system method to calculate the unloading risk data set. Based on this dataset, an offloading risk assessment model is proposed by combining the BP neural network algorithm. After training and iteration, the model has an accuracy rate of 95%, and the loss value is maintained at a very low level of 0.099, which shows an excellent risk assessment capability.
