Water Science and Technology (Mar 2023)

A comparison of struvite precipitation thermodynamics and kinetics modelling techniques

  • Leynard Natividad-Marin,
  • Max William Burns,
  • Phil Schneider

DOI
https://doi.org/10.2166/wst.2023.061
Journal volume & issue
Vol. 87, no. 6
pp. 1393 – 1422

Abstract

Read online

Solution thermodynamics and kinetic modelling applied to struvite crystallisation–precipitation were reviewed from diverse references to determine proximity between predicted and cited experimental measurements. These simulations show the expected variability range of struvite saturation calculation when only limited solution compositional information is given, showing acceptable agreement between predicted and experimental struvite mass. This work also compares results from struvite crystallisation kinetic studies on liquid phase species depletion, crystallisation induction time, primary nucleation, secondary nucleation, crystal growth, and crystal aggregation. Large inconsistencies between reported kinetics were observed in many scenarios. Variations in species depletion models highlighted that they are only suitably applied to the specific system from which they were regressed. Spontaneous primary nucleation was predicted to occur in the range of SI = 0.237–0.8. Predicted primary nucleation rates vary over at least 10 orders of magnitude (depending on supersaturation) because of uncertainties in interfacial tension and maximum achievable nucleation rate. Secondary nucleation rates are more agreeable, varying over approximately two orders of magnitude. Growth rates varied over five orders of magnitude due to variations in experimental conditions. Aggregation rates are not thoroughly examined enough to make any inferences. HIGHLIGHTS Assessment of model structure and predictive capability for struvite thermodynamics in saturated systems.; Applicability of simulations to design model validation using real effluent for struvite precipitation.; Primary nucleation models are highly variable while secondary nucleation models are more consistent.; Increased model complexity is required for improved struvite growth rate prediction accuracy; Growth rate orders cannot be accurately used to infer crystal growth mechanism.;

Keywords