VRLA battery float charge : analysis and optimisation.
Thesis DisciplineElectrical Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
This thesis investigates the float charge operation of Valve Regulated Lead Acid (VRLA) batteries in standby power applications. While the telecommunication standby power system is the targeted application, the results are applicable in any situation where VRLA batteries are subjected to long periods of float charge. The goals of float charge are identified and a test and analysis procedure is developed to provide a means of assessing the effectiveness of the applied float charge. The primary goal of float charge is to counteract the natural self-discharge of the battery and indefinitely maintain it in a fully charged state. A secondary goal of float charge is to maximise the life of the battery. This is achieved by ensuring that ageing mechanisms such as positive grid corrosion and gas venting are maintained at minimum levels. The problems associated with conventional float charge control are investigated and, in particular, the electrode imbalance problems associated with some long life VRLA cells are detailed. These electrode imbalance problems can result in the cell suffering a gradual discharge of the negative electrode while the cell appears healthy and on a float charge. This ultimately results in reduced cell capacity and is identified as a major cause of the premature failure of long life VRLA cells. Float charge analysis and the subsequent optimisation relies heavily on knowledge of the polarisation distribution between the positive and negative electrodes within a cell. Conventionally, this is determined with the aid of a reference electrode, although such testing is only possible in a well-controlled laboratory environment. By modelling the steady state and transient features of both the positive and negative electrodes, a test and analysis procedure is developed to estimate the polarisation distribution within a conventional 2 V VRLA cell, in effect creating a virtual reference electrode. The developed procedure exploits differences in the transient response of each electrode to estimate their polarisations at the applied float voltage. The polarisation estimations are typically accurate to within 10 mV, the window of polarisation relating to minimal positive grid corrosion is approximately 40 mV wide, and the total polarisation applied to a VRLA cell is around 130 mV. The test requires only a very low rate constant current discharge, and cell terminal voltage measurements. This test may be automated and applied to cells in field service, and provides the necessary measure to gauge float charge optimisation. The developed test is able to verify that a cell on float charge is indeed fully charged, and assists in determining the optimal float voltage for maximum cell life.