Co-complexation effects during incineration bottom ash leaching via comparison of measurements and geochemical modeling

Under the proper setting of boundaries, leaching behaviors of incineration bottom ashes (IBA) may be predicted using geochemical models. The extent of agreement between measurements and the prediction is often affected by the model development. On the other hand, the inconsistency may also be caused by the complexity of the system for modeling. As such, discrepancies between measurements and geochemical modeling on IBA leaching would indicate uncovered yet significant mechanisms which may not be fully understood by the existing setting scenarios. pH-static leaching tests of IBA were performed, followed by simulations with two geochemical models. Simulations based on the surface area calculated with all leachable minerals provided a better matching with experimental data as compared to that calculated with surface area of Fe/Al oxides only. Yet, certain elements remained highly overestimated (up to 3 orders of magnitude, e.g. Cd, Ni, Sr, Zn, Fe, Al and Mg) or underestimated (e.g. Cr) in alkaline pH range. Significant co-complexation during leaching may be susceptible and responsible for the differences. Multivariate factor analysis testified that pH, Fe and Al in acidic environment tended to be clustered and highly loaded under the dominant correlation factor (accounting for 33.3-49.0% of variance). Contrastively, pH, Ca and cationic strength in alkaline environment tended to be clustered and highly loaded under the dominant correlation factor (accounting for 39.9-43.3% of variance). In both cases, cocomplexation was evidenced. As the natural pHs of IBA usually fall within the high alkaline range, the comparison study offered quantitative discrimination of leaching potential between measurements and simulation, which would be crucial for decision making during IBA applications. In light of limited leaching potential of IBA under the alkaline pH range, its application with cement (with high alkalinity) seems to be one of the best options. (C) 2018 Elsevier Ltd. All rights reserved.