Fault Rates Alpha-Tolerance of Monte Carlo Simulation (FRAT-MCS) for Reliability Performance of Power System Network

This paper presents a Fault Rates Alpha-Tolerance Monte Carlo Simulation (FRAT-MCS) method for evaluating reliability in electrical distribution networks. Conventional Monte Carlo Simulation can produce results that differ significantly from actual utility fault data, leading to unreliable estimates of system performance. To address this, FRAT-MCS introduces upper and lower tolerance bounds on the number of interruptions for each network component. This ensures that simulated fault counts remain close to expected values based on real fault rates. The method also uses time-varying load profiles and daily fault probability curves. These features allow the simulation to reflect how demand and fault likelihood change over 24 h. The approach is tested on IEEE distribution test systems and compared with two other Monte Carlo methods. Results show that FRAT-MCS produces reliability indices with smaller standard deviations and tighter confidence intervals. This indicates improved precision and stronger repeatability. The method avoids generating unrealistic extreme interruption patterns while still capturing meaningful year-to-year variability. Statistical tests confirm that FRAT-MCS delivers stable and consistent outcomes across multiple simulation runs. These characteristics make FRAT-MCS suitable for network planning, regulatory reporting, and decision-making where reliable and reproducible reliability estimates are essential.