Nuclear-based hydrogen generation is a promising way to supply hydrogen for this large market in the future. This thesis focuses on one of the most promising methods, a thermochemical Cu-Cl cycle, which is currently under development by UOIT, Atomic Energy of Canada Limited (AECL) and the Argonne National Laboratory (ANL). The safety issues of the Cu-Cl cycle are addressed in this thesis. An investigation of major accident scenarios shows that potential tragedies can be avoided with effective risk analysis and safety management programs. As a powerful and systematic tool, fault tree analysis (FTA) is adapted to the particular needs of the Cu-Cl system. This thesis develops a new method that combines FTA with a reliability analysis tool, multi-state system (MSS), to improve the accuracy of FTA and also improve system reliability. / UOIT