Statistical Comparison of Two Studies on Trace-Element Composition of Coal-Ash Leachates. Final Report

摘要

Data from two studies of the measurement of trace-element concentrations in ash from coal-fired power plants were reanalyzed using a common statistical methodology. Both studies involved replication of the sample preparation (extraction) and analytical processes at multiple laboratories. Variance components analysis was used to measure the contribution of inter- and intra-laboratory extraction and analysis variation to the overall precision of the concentration measurements. The data was transformed to the logarithmic scale, screened for outliers, and appropriate handling of detection-limit values was implemented prior to the statistical analysis. Inter-laboratory effects were found to be generally larger than intra-laboratory effects, though all sources of variation had a noticeable effect on the precision of the measurement. In both studies, inter-laboratory analysis variation was found to have a large effect, and in one of the studies, inter-laboratory extraction variation was found to have a large effect. The differences between the two studies may be due to differences in the reference ash used, to differences in the amount of inter-laboratory quality control on the extraction and analysis procedures, or to differences in the element-concentration levels in the samples used in the two studies. (ERA citation 08:045360)