Determining Concentration Inhomogeneity Inside Liquid Holding Tanks

Incorporating concentration measurements of special nuclear material (SNM) in a chemical processing accountability program requires uncertainty data for three types of errors: mixing, sampling, and analytical. Usually, the analytical measurement uncertainties—both random and systematic—are routinely determined for the laboratory instruments and procedures being used to measure the concentration of withdrawn liquid samples. However, little data currently exists to determine how representative the withdrawn liquid sample is in relation to the true average bulk concentration within the tank itself. This sample representativeness depends upon uncertainties associated with the tank mixing and sampling procedures. Both types of uncertainties are investigated. Mixing uncertainty or inhomogeneity has been found to be primarily a fluid mechanics problem while sampling uncertainty involves mostly chemistry. Experimental results are presented along with the early developments of a scaling analysis for future process design and accountability work. Tests were performed on full size pencil, double pencil, and annular tanks using saltwater solutions to achieve various levels of density stratifications within the tanks analogous to nitric acids solutions of different normalities.