This paper investigates the short-term extreme value distribution of the tension on mooring lines and the reliability of grid mooring system for fish cage under ultimate limit state. The generalized probability density evolution method (GPDEM) is employed to determine the evolution of the probability information of mooring tension. The short-term extreme tension distribution of mooring lines is evaluated through the combination of the GPDEM and the virtual stochastic process model. And the reliability analysis under ultimate limit state is conducted based on the extreme tension distribution of mooring components in grid mooring system. In addition, the equivalent extreme-value event method is adopted to perform the reliability analysis of grid mooring system for fish cage. The results indicate that the equivalent extreme-value event method has a significant advantage over the narrow bound approach in accurately predicting the system reliability and a large uncertainty of the random allowable strength largely affects the system reliability of grid mooring system. Therefore, it is of vital importance to perform the uncertainty quantification of the allowable strength for accurately conducting the system reliability analysis in order to ensure the safety of offshore net cage structures.
