A Statistical Study of Sources of Variation in Primary Battery Testing

Abstract

An analysis of variance (mixed model) experimental design was employed to quantify sources of variation in primary battery testing procedures for lithium/fluorinated carbon (CF//x) systems. Factors investigated were mix repeatability, mix uniformity, electrode fabrication method, and cell assembly and testing. Parameters measured were percent utilization and closed-circuit voltage (CCV) at various depths of discharge (DOD). The standard deviations (for a random cell from a random blend on a random day) for percent utilization and CCV at 50% DOD were 4. 2 percentage points (0. 49 relative standard deviation) and 16 mv (0. 006 relative standard deviation), respectively, indicating that the test method was adequate for evaluation of the performance of different CF//x samples. No significant difference due to electrode fabrication method in any of the measured parameters was observed.