Comparisons of Algorithms for End-of-Life Detection of Lead-Acid Batteries Used in Off-Grid PV System

For remote PV plants, having a way to detect battery's end-of-life with some weeks in advance is a need. There are several algorithms for such detection. However, how they could work at field plant is not clear, neither is the performance comparison among them. The paper presents a direct comparison among some algorithms for lead-acid batteries, suited for field use. They were tested against sets of data collected with batteries' accelerated ageing tests. Algorithms are: rate of voltage change, internal impedance, direct entropy and modified charge factor. Direct entropy has proven being most robust of them.