Comparison of uranyl nitrate solution modeling methods

In assessing the criticality of actinide solutions, highly accurate determination of the solution composition is essential. This is fairly simple for binary solutions where ample data are available. However, it becomes much more difficult if solutions contain many components and are at temperatures other than 25°C. Such applications have become increasingly important in evaluating evaporators, dissolvers, and elevated-temperature waste tanks. These may contain fissionable solutions involving mixed actinides and even multiple anions. A comparison has been made between three solution modeling techniques, including the resulting atom densities and the calculated k _eff results using SCALE. Three experiments involving uranyl nitrate solutions with high 233U content inside varying diameter cylinders were used as the basis for comparison. This comparison reveals a significant difference in both atom densities and calculated k _eff results depending upon which solution modeling technique is used. In some cases, the difference in calculated k ^eff for a given experiment is as great as 1 percent. The ARH-600 method produces a negative bias in k _eff when compared against the Pitzer method. The ARH-600 method produces a negative bias in k _efffor two experiments and a positive bias in k _eff for the other experiment when compared against the Söhnel and Novotn_ method. The Söhnel and Novotn method also displays a negative bias in k _eff for two experiments and a positive bias in k _eff for the other experiment when compared to the Pitzer method. All three methods exhibit minor differences in atom densities for the uranium isotopes. Differences in atom densities for the three solution methods are primarily apparent for hydrogen, nitrogen, and oxygen.