Most of the Wireless sensors today are powered using lithium batteries. The need to detect the end of life for these batteries has increased in the recent past due to a large manufacturing variation between suppliers and storing practices. It has become a challenge to explore the ways to access the health of these batteries in field due to the difficulty in measuring the internal state variables of the battery and also due to its inaccessibility. Failures of batteries in the field not only result in inconvenience and reduced availability of the wireless sensor, but also risk catastrophic consequences due to the unavailability of sensors to detect the anomaly. As a result, the current research work focuses on evaluating the battery end of life through a laboratory testing that could reproduce the field environment. Using the measured state variables an attempt is made to predict the battery health before it is introduced to the field. The research also entails developing a bulk capacitance system that can act as secondary charge storage to help increase the primary battery life during peak current draw. Using a tree based classifier a prediction methodology is developed considering both the battery and bulk capacitance parameters to estimate the overall battery health or end of life.