Adrian E. Flood

Model-driven experimental evaluation of struvite nucleation, growth and aggregation kinetics

Nutrient stewardship is emerging as an issue of global importance, which will drive the development of nutrient recovery in the near to medium future. This will impact wastewater treatment practices, environmental protection, sustainable agriculture and global food security. A modelling framework for precipitation-based nutrient recovery systems has been developed, incorporating non-ideal solution thermodynamics, a dynamic mass balance and a dynamic population balance to track the development of the precipitating particles. The mechanisms of crystal nucleation and growth and, importantly, aggregation are considered. A novel approach to the population balance embeds the nucleation rate into the model, enabling direct regression of its kinetic parameters. The case study chosen for the modelling framework is that of struvite precipitation, given its wide interest and commercial promise as one possible nutrient recovery pathway. Power law kinetic parameters for nucleation, crystal growth and particle aggregation rates were regressed from an ensemble data set generated from 14 laboratory seeded batch experiments using synthetic solutions. These experiments were highly repeatable, giving confidence to the regressed parameter values. The model successfully describes the dynamic responses of solution pH, the evolving particle size distribution subject to nucleation, growth and aggregation effects and the aqueous magnesium concentration in the liquid phase. The proposed modelling framework could well be extended to other, more complex systems, leading to an improved understanding and commensurately greater confidence in the design, operation and optimisation of large-scale nutrient recovery processes from complex effluents.

Thermodynamic parameters and counterion binding to the micelle in binary anionic surfactant systems

Competitive counterion binding of sodium and calcium to micelles, and mixed micellization have been investigated in the systems sodium dodecylsulfate (NaDS)/sodium decylsulfate (NaDeS) and NaDS/sodium 4-octylbenzenesulfonate (NaOBS) in order to accurately model the activity of the relevant species in solution. The critical micelle concentration (CMC) and equilibrium micelle compositions of mixtures of these anionic surfactants, which is necessary for determining fractional counterion binding measurements, is thermodynamically modeled by regular solution theory. The mixed micelle is ideal (the regular solution parameter β(M)=0) for the NaDS/NaOBS system, while the mixed micelle for NaDS/NaDeS has β(M)=-1.05 indicating a slight synergistic interaction. Counterion binding of sodium to the micelle is influenced by the calcium ion concentration, and vice versa. However, the total degree of counterion binding is essentially constant at approximately 0.65 charge negation at the micelles surface. The counterion binding coefficients can be quantitatively modeled using a simple equilibrium model relating concentrations of bound and unbound counterions.

Feedback between crystal growth rates and surface roughness

This article analyzes a collection of recent research on a recently discovered feedback mechanism between the crystal growth rate and the surface features of the crystal, and the ramifications the mechanism has on crystallization in general and also on the design and analysis of industrial crystallizers. It has been found that growth under high supersaturations degrades the crystal surface, causing a roughening that is probably due to imperfect incorporation of growth clusters into the surface of the crystal. The effect becomes more pronounced under higher growth rate conditions, and for higher residence times under such conditions. The mechanism only occurs if the supersaturation is increased above a critical level known as the macroscopic roughening transition, although this level is typically quite low; for sucrose it has been measured as a relative supersaturation between 2.5 and 3.9% which is in the range used in many industrial crystallizations. The mechanism appears to be associated with the surface energy of the crystal, with the surface of high surface energy crystals being degraded at lower values of the supersaturation, and such crystals also have larger reductions in growth rate after the roughening has occurred. The mechanism also appears related to growth rate dispersion (GRD), since it gives a mechanism for variation in growth rates in batches of crystals, and also because GRD also appears to be more significant in species having a higher value of the surface energy. The mechanism also causes an increase in the impurity incorporation in the crystal, thus leading to reduced product crystal purities. The mechanism for the impurity incorporation is probably due to enhanced adsorption of impurity molecules due to large crystal surface areas (due to the roughening) and also larger numbers of adsorption sites. The mechanism also complicates measurement of crystal growth kinetics, and thus makes efficient design of industrial crystallization units more difficult.

NUMERICAL SIMULATION AND ANALYSIS OF FLOW IN A DTB CRYSTALLIZER

This research numerically simulates the two-phase (liquid and vapor) flow in a 1 m3 draft tube baffle (DTB) crystallizer with fines removal streams. The computational fluid dynamics (CFD) commercial software ANSYS CFX-10.0 was employed to perform 3-D simulation using the finite volume method with an unstructured mesh topology. The influence of hydrodynamics in the crystallizer, as characterized by the momentum source strength and fines removal flow, on the flow characteristics and the classification of crystals are investigated. The results showed the liquid flow is fully uniform in the main body of the crystallizer studied for momentum sources larger than or equal to 19.63 kg · m/s2. The uniformity of the suspension will strongly affect the product crystal size distribution. Momentum source strengths and fines removal flow rates also have a significant effect on the fines removal cut-size due to varying up-flow velocities in the fines removal section, altering the size at which particles are carried out in the fines removal stream. The CFD predictions are compared with the experimental results from the literature and can be used for the optimization of commercial-scale DTB crystallizer design.