In power distribution systems, one of the most dangerous events is the occurrence of lightning surges. Lightning surges directly impact overhead distribution lines and then propagate to other vital component of the network such as transformers, underground cables etc. Due to the extended calculation process of random nature of the problem, the use of Artificial Intelligence Techniques for failure risk prediction is highly advantageous as it reduces effort and saves time. In this paper AI techniques like neural network(NN), fuzzy logic(FL) and neuro fuzzy techniques(NF) along with the surge arrester ratings are used to predict the risk of power system failure. Simulated results evince the superiority of the Neuro fuzzy techniques