A Stochastic Weather Dependent Reliability Model for Distribution Systems

In power system planning and operation, accurate assessment of reliability worth is essential for making informed decisions. One common simplification when modeling power system reliability is assuming constant failure rates and non time-varying restoration times. However, historical outages show differently; failure rates and restoration times for especially overhead lines are dependent upon time-varying factors as, for instance, weather conditions. When modeling this time dependence a two or three-state weather model is often used. The reliability model proposed in this paper does in contrast use the stochastic nature of the severe weather intensity and duration to model variations in failure rate and restoration time. Further, the model also considers when severe weather is likely to occur during the year by using a non-homogeneous Poisson process (NHPP). A time-sequential Monte Carlo technique is applied to a radial distribution system. By combining the proposed reliability model with a time-dependent interruption cost model, the effect of the inclusion of time-varying failure rates and restoration times is investigated and found to be of importance when assessing reliability worth.