Design of a static mixer reactor for copper recovery from waste streams

More Info
expand_more

Abstract

The main goal of the project was the development of a plug flow reactor for the reduction of heavy metals (Cu2+) from industrial waste streams. Potential application of the reduction process inside The Netherlands lies in the IC and galvanic industry, where small waste streams containing aqueous copper exist. Outside The Netherlands, the process could be applicable in the mining industry,e.g. in Chili or South Africa. The copper is reduced in the form of particles by soluble carbohydrates, which provide the electrons for the precipitation. The carbohydrates may originate from another waste stream, which can be found in the food or wood industry. After hydrolysis, these carbohydrates can be applied as reductor. Furthermore, the carbohydrates are degraded, which lowers their carbon oxygen demand and cleans the waste streams biologically. This way, the two waste streams are cleaned simultaneously and a valuable end product in the form of copper particles is recovered. The main focus of this thesis is on the application of the Kenics static mixer in a pipe reactor, in order to achieve plug ow conditions in such a reactor. The static mixer is used to control the residence time of the particles, and to mix the chemical species in the reactor. The key question is under which conditions the application of the static mixer leads to a (more) narrow particle size distribution. A narrow size distribution of the particles is an important aspect, since it enhances the economical value of the end product. To answer this question, the Kenics static mixer is studied in detail both numerically and experimentally. The flow in the Kenics static mixer has been investigated both numerically and experimentally in the range of Re=100 1000. It was found that at Re=300 the ow becomes unsteady. Two numerical methods, the Lattice Boltzmann (LB) method and the Finite Volume method (FLUENT) were compared and used to simulate the flow. The LB method proved to be a relatively fast and cheap (in terms of memory) alternative for the simulation of the transient flow in the Kenics static mixer at Re>300. Furthermore, the flow field and dynamic behaviour were validated by means of LDA experiments. The transient behaviour observed was explained by studying the dynamics of the vortices in the flow. The transition to unsteady flow takes place, when the vortices start stretching out over an entire mixing element and start creating a disturbance in the flow entering the next mixing element, which subsequently triggers the unsteady behaviour. To investigate the behaviour of particles in the static mixer, a Particle Tracking (PT) code is developed and linked to and embedded into the LB code. The particle tracking code is based upon the equation of Maxey and Riley (1983) to which the modified lift force (Saffman (1965, 1968)) is added. Furthermore, a growth model for the particles is added to the PT code. The particle growth is based upon the diffusion of Cu2+ to the surface of the particle. The Cu2+ concentration is solved with a standard finite volume code, which solves the convection-diffusion equation with a sink term. The sink term is directly linked to the growth of the particles present in the finite volume cell. The chemical parameters due to Van der Weijden et al. (2002a) are used as input for the growth model, where it is assumed that the diffusion of Cu2+ is the rate limiting factor. The results indicate that two important design parameters for the Kenics static mixer reactor are the Reynolds number, which is a measure of the flow regime, and the St/Fr ratio, which is a measure of the settling rate of the particles. The two numbers determine to a large extent the mixing, settling and residence time of the particles. Ideally, the particles are uniformly distributed and have an uniform residence time distribution (plug flow). It was found that these conditions were best matched at a low St/Fr ratio (St/Fr < 1) and at either a low or a high Reynolds number (Re < 20 or Re > 200). In a horizontal reactor, settling of the particles poses a problem that is directly related to the St/Fr ratio. It was found that in order to keep the majority of the particles in suspension the St/Fr ratio should be small and the Reynolds number high (St/Fr < 0.01 and Re > 500). Alternatively, the reactor can be placed vertical. If the flow direction is downward in such a reactor, no problems regarding the settling of particles occur, which removes the limit on the St/Fr ratio. However, there remains a limit regarding the mixing of the particles. When the St/Fr ratio is high (> 1), particles collide with the mixing element, which leads to accumulation of the particles near the mixing elements. It was investigated what the in uence of this ill-mixing of the particles was on the particle size distribution. For that purpose, simulations were carried out of growing copper particles in a vertical KenicsTM static mixer reactor. It was found that the particle size distribution is wide, when the particles are not mixed effectively. Therefore, a vertical reactor is also limited by the St/Fr ratio (St/Fr < 1), when a high quality end product is required, i.e. particles with a narrow size distribution. The results of the chemical (autoclave) investigations are combined with the numerical results, to propose a design for a continuous (plug flow) reactor. A one-dimensional model is used to predict the reduction of Cu2+ in three reactor configurations (batch, horizontal plug flow reactor and vertical plug flow reactor). Experiments in a glass-lined autoclave were used to test the model and to obtain the model parameters. The model is used to predict the (mechanical) energy consumption per kg recovered copper. Furthermore, the total energy demand of the process (heating + pumping/stirring) was evaluated for different reactor types and compared to electro-winning being the conventional method of copper recovery. It was found that heating the liquid towards the set temperature is the main energy consumer. Based upon its energy demands, the applicability of the reactor is assessed for industrial waste treatment and the mining industry. It was found that the vertical plug flow reactor can be an attractive alternative for electrolysis, when the stream has a high Cu2+ concentration or when the stream is contaminated with organic material. It should be noted that the vertical reactor was explicitly designed for the treatment of small waste streams that exist in the Netherlands. For processing the large streams that exist in the mining industry, the throughput of the vertical reactor is too low. This limitation can be overcome by placing different vertical reactors in parallel to accommodate a large throughput. However, the use of another type of static mixer might extend the feasibility of the vertical reactor towards a higher throughput. The design of such a 'large' vertical reactor can be an interesting topic for future investigations.