Error analysis in the measurement of current efficiency in Hall-Héroult cells - Part II: Application and results

Upon cursory consideration, the determination of current efficiency for a Hall-Héroult cell appears to be a deceptively simple task. After all, what is required but the measurement of metal production as a function of the cumulative ampere-hours of current flow? However, when attempts are made to determine the current efficiency of a single Hall-Héroult cell over a relatively short elapsed time (less than 48 hours), the data rarely show accuracy, precision, or reproducibility. This paper discusses the statistical error, which is associated with the determination of current efficiency as a function of the various field and analytical measurements that are required for its calculation. A general statistical model is developed that can be applied to a number of different measurement techniques and strategies, and to any type and size of Hall-Héroult cell. The model is used to show how the error associated with current efficiency measurement is influenced by factors such as cell size, which is a function of design amperage and metal pool inventory, and elapsed time. This paper is divided into two parts. Part I describes the fundamental aspects of current efficiency measurement and introduces the law of propagation of errors. Part II applies this law to the estimation of error in current efficiency measurement using several practical examples.