A statistical analysis has been made on data obtained in the study of the leaching process for the American Society for Testing and Materials special fly ash specimen using distilled water as the extractant. Two laboratories made extractions under a variety of conditions, and after acid stabilization, the leachates were analyzed by a third laboratory using inductively-coupled plasma analysis. Fifteen elements were studied. Three types of experimental variables were examined: (1) the extractor (60 and 180 cpm shaker tables, and the National Bureau of Standards (NBS) mixer); (2) the liquid/solid ratio (4/1, 10/1, and 20/1); and (3) the time (24 and 48 h).

The 60 cpm shaker table at the 4/1 liquid/solid ratio and at 24 h showed the greatest variability of results; the NBS mixer showed the smallest variability. The NBS mixer and the 180 cpm shaker table exhibited very similar extraction efficiencies and were, for a number of elements, noticeably different from that of the 60 cpm shaker table.

The data were examined in terms of three plausible extraction models. Many of the elements studied fit the model that practically all of the elemental material capable of being extracted is extracted under the test conditions, and that for the higher liquid/solid ratios this material is simply diluted. This results in the concentration being inversely proportional to the liquid/solid ratio. Data for the remaining elements appear to be in better accordance with a leaching model based on interrelated chemical equilibria. We observe, for example, that when the aluminum concentration decreases, the silicon concentration increases. This effect appears to be so pronounced that for the cases of the NBS mixer and the 180 cpm extractor, the silicon concentration actually goes up for the higher liquid/solid dilution ratios. A third plausible leaching model which is based on the concept of a constant elemental saturation under all experimental conditions generally was not found to be supported by the data.