The battery is an essential component in providing continuous electricity supply using renewable energy sources. It can be found in many daily applications, such as in the telecommunication system, radio microwave system, emergency lighting, the backup system of power plants, even in a photovoltaic system. It is often used as the backup source in case of a failure in the main supply system. The duration of how long the battery can still supply energy to loads without being charged is defined as the battery autonomy day. If during its daily utilization the battery often exceeds its autonomy day, it can result in the deterioration of the battery lifetime. It produces the deviation of the battery lifetime specification which has been previously determined by the manufacturer. This paper presents the results of battery lifetime prediction at a base-transceiver station (BTS) of Telkomsel Company in Indonesia. It has two main purposes which are to evaluate the policy of autonomy day and to predict the remaining lifetime of the battery before reaching its time limit. The obtained results show that there have been some alterations from the batteries' former policy of autonomy day, from 72 hours to 43.03 hours and 43.26 hours for both existing batteries respectively with considered depth-of-discharge (DOD) of 20%. By using a linear data curve fitting, the results of calculation and analysis indicate that the remaining useful lifetime of both batteries were 5.72 years and 5.77 years. Another approach using an exponential data curve fitting resulted in the remaining lifetime of 7.12 years and 7.16 years for both batteries respectively.