Reliability, availability and maintainability (RAM) study of a power generation plant has been conducted using Monte Carlo simulation. The simulation is based on reliability block diagram (RBD) of power generation process, mean time between failure (MTBF), and mean time to repair (MTTR) data. From the simulation, the system 25-year mean availability is predicted at 93%. An alternative method to reliability block diagram is the Stochastic Petri Nets (SPN) model. SPN is one of the established mathematical language tools to conduct various types of concurrency, synchronization and sequencing in distributed systems. This paper presents RAM study of a power generation plant using SPN modeling. In this model, the maintenance and repair teams are considered always available at any time regardless of how many components fail at any particular time. The logistic time for supplying required spare parts is not considered, and it is assumed that there is no delay in parts supply. The result shows that the mean availability of the system from both techniques are in good agreement. In fact, SPN modeling demonstrates a smoother exponential curve in analyzing the mean availability of the system compared to the RBD simulation. By assuming MTBF and MTTR as constant, predicted number of failures for all components are between 13 to 84 times over 25 years of operation.