Household electricity consumption is a direct contributor to household expenses. Electricity acts as a backbone for a strong economy . The rise in the energy consumption is clearly observed in this past decade, and so is the rise in the need for energy efficiency and conservation . Monitoring power consumption by using various devices and instruments is on the rise; however a smart environment scenario needs more than just real-time monitoring. The need for identifying abnormal power consumption is clearly present. In this paper, we introduce our work on building novel outlier detection algorithms which uses statistical techniques to identify outliers and anomalies in power datasets collected in smart environments. We also experiment clustering techniques on the same dataset and report the results found.