The effects of lightning strike to transmission and distribution systems are numerous and unfavorable. Just imagine the cost and havoc if a giant telecommunications company is shut down for an hour or day as a result of devices damage or a petrochemical plant catches fires due to lightning strike. Hence the needs to protect power apparatus from overvoltage surge are imperative. In this study an adaptive risk analysis & management (ARAM) based on artificial neural networks is proposed to analyze and proactively control the overvoltage at power substation due to lightning strike. General TermsMachine learning classification problem and System design.