Stochastic Petri nets for reliability assessment of power generating systems with operating considerations

This paper presents an alternative reliability assessment technique that takes advantage of the strengths of present analytical and Monte Carlo reliability methods: speed and flexibility. The new method is based on stochastic Petri net techniques. Similar to Monte Carlo simulations the new method recognizes important operating considerations such as unit hot and cold startup delays, startup failures distinct from running failures, annual planned outages, and reserve requirements. The method calculates several performance measures such as expected loss of load, expected unserved energy, expected duration of loss of load, expected frequency of loss of load, etc. Results from case studies using the EPRI system are presented. Comparison of results with those from conventional simulations show the new method to be computationally faster and less tedious