The Flow of Solution and Crystals in a Draft Tube Crystalliser Investigated Using Positron Emission Particle Tracking

G.S.M. van Beuzekom1,2, M.D. de Jong1,3, Z.I. Kolar2, H.J.M. Kramer3, B. Scarlett1

1 Particle Technology Group, Faculty of Applied Sciences, Delft University of Technology, Julianalaan 136, 2628 BL, Delft, the Netherlands, beuzekom@dds.nl

2 Interfaculty Reactor Institute, Delft University of Technology, Delft, the Netherlands

3 Laboratory for Process Equipment, Delft University of Technology, Delft, the Netherlands

ABSTRACT

Due to the complex multiphase nature and hydrodynamics, industrial scale crystallisation processes are not well understood yet. Previous research led to the development of Computational Fluid Dynamics (CFD) models describing the flow patterns in such equipment. These models have not yet been validated. Therefore the goal of the current research is to validate those models using Positron Emission Particle Tracking (PEPT).

PEPT is a radiotracer technique, which uses a single positron emitting source. A positron annihilates an electron, resulting in two 511 keV D-quanta. These D-quanta move away from each other in a back- to-back fashion. They then can be detected coincidentally by two opposite facing position-sensitive scintillation cameras, resulting in a reconstruction line. The 3D tracer location is established by computing the approximate intersection of a set of these lines. If sufficient locations can be determined within a predetermined interval, it is possible to map 3D movements of the tracer. It is also possible to compute tracer residence times and mass fluxes. The advantage of PEPT over other measurement methods is that radioactivity is capable of penetrating the dense and opaque slurry flow of crystallisers.

The subject of this study is a 22 litre solid-liquid stirred Draft Tube Crystalliser (DTC). To avoid disturbance of flow measurements by crystal growth phenomena, fake crystals (glass beads) and plane water were used. The crystals display a Crystal Size Distribution (CSD), with a mean crystal size of 650 om.

With two different types of radiotracers the flow behaviour of the glass beads (crystal tracer) and the flow behaviour of the water in absence of crystals (solution tracer) have been investigated. All radiotracers contain 64Cu as positron source (half life 12.7 h). Each tracer particle starts with an activity of 20-22 MBq, and was used until its activity dropped below 10-12 MBq; this takes approximately 14 h. The optimum tracking activity lies at 14.8 MBq.

The resulting flow patterns all correspond very well with those predicted by CFD. PEPT was able to confirm the existence of a vortex over the stirrer, which also appeared in the CFD models and was always thought to be a numerical artefact. PEPT was also able to quantify differences in pure liquid and suspension flow behaviour. In view of the unexpected results, it is concluded that PEPT is a powerful and non-invasive technique for quantitative and qualitative determination of the flow of solution and crystals in bench scale crystallisation equipment.

Keywords PEPT, radiotracer, validation, multiphase flow