Linh T.T. Vu

Modelling Cobalt Solvent Extraction using Aspen Custom Modeler

Abstract The cobalt solvent extraction system using Cyanex 272, a phosphinic acid based extractant, has been modelled using the Aspen Custom Modeler mathematical modelling software. The principle advantage of this method is that the model can easily be imported into Aspen Plus and run as part of an integrated flowsheet containing other unit operations. The cobalt solvent extraction circuit operates on a counter-current basis, with the barren organic entering the final stage and the aqueous feed entering at the first stage. Since the metal extraction efficiencies were dependent on the conditions of the outlet streams, a solver must be selected to simultaneously solve a set of algebraic nonlinear model equations. Initial sensitivity analysis for a single stage Aspen Custom Modeler model has shown that increasing pH or the organic to aqueous (O:A) ratio significantly increases individual metal extraction efficiencies. To achieve the ultimate aim of maximising cobalt extraction while minimising magnesium and nickel co- extraction and reagent consumption, an economic objective function has been formulated within the optimisation problem to solve for the optimum pH setpoint and O:A ratio. The optimised single stage results indicate operating at pH 4.5 and O:A of 0. 78 to achieve 95% cobalt extraction, while limiting nickel extraction to

Development of an integrated model for cobalt solvent extraction using Cyanex 272

A model of metal extraction based on pH isotherms was generated and applied to a cobalt solvent extraction (SX) circuit. Cyanex 272 (bis-(2,4,4-trimethylpentyl) phosphinic acid) was used as the organic extractant due to its selectivity for cobalt over nickel in the extraction process. Experiments were conducted for cobalt, nickel and magnesium extraction, with the latter two representing impurity elements in Co SX. The methods for determining metal extraction incorporated the effects of temperature, solution composition and pH on the equilibrium constant k, and hence on the overall extraction extent. This information was applied to a multi-stage mixer-settler model consisting of integrated extraction units. The initial mathematical model for cobalt, which was built in Matlab can be further developed to include the impurity elements and incorporate the scrubbing and stripping units. Future work will focus on using the model for process optimisation.

Steady state optimization of design and operation of desalination systems using Aspen Custom Modeler

In this paper Multistage flash (MSF), Reverse Osmosis (RO) and hybrid MSF/RO desalination systems are optimized. A superstructure is set up for analyzing various process configurations in a single flowsheet. Detailed steady state models for MSF and RO sections of the superstructure are developed incorporating comprehensive physio-chemical properties and design characteristics. The model for the hybrid system combines individual models of MSF and RO systems and additional separators and mixers. The optimization variables consist of the operating and design variables, and the objective function is developed on the basis of economic and technical performance indicators. Primary results show that the hybridization of MSF and RO systems sharing common intake-outfall facilities presents lower cost with the additional benefits of higher overall recovery than MSF system and higher product quality than RO system. The sensitivity analysis of the cost parameters is performed to realize their effects on the selection of process configuration.

Superstructure Development, Simulation and Optimization of Desalination Systems using Aspen Custom Modeler

In this paper, Multistage Flash (MSF), Reverse Osmosis (RO) and hybrid MSF/RO systems are modelled and simulated in Aspen Custom Modeler (ACM) V.8.4. Both steady state models include material balances; but RO focusses more on pressure differences, while MSF relies on energy balances. The model of a hybrid superstructure is developed by combining individual models of MSF and RO systems. The hybrid superstructure is developed in an optimization framework to analyse and optimize various process configurations. Operating conditions such as temperatures, pressures and process design variables such as the number of membrane elements in a pressure vessel and the number of stages in an MSF system are used as optimization variables to minimise the economic objective function. Primary results show that the hybrid MSF/RO systems sharing common intake-outfall facilities have lower annual intake cost and can be considered in the areas where high overall product recovery and high product quality is required.

Dynamic optimization of desalination system designs using Aspen Custom Modeler

The economical and sustainable desalination processes are of supreme importance. The two desalination processes leading the market are the Multistage Flash (MSF) and Reverse Osmosis (RO) systems. New configurations by combining the MSF and RO in a hybrid system have also been studied. The hybrid desalination systems are believed to provide better water quality and lower power demand. However, there is very limited work around modeling and optimization of these systems. Moreover, no attempt has been made to study the transient behaviour of these systems. Thus further insight into these processes is required before implementing them on an industrial scale. In this paper, dynamic optimization is conducted to select the best configuration based on the economic objective function. The formulated dynamic optimization problem determines the optimal process configuration and process variables for different feed water concentrations. The optimization problem is implemented using a superstructure and the open loop dynamic models based on the first principles. The results of the study show that the cost reduces over time as the optimizer trade-offs between the capacities of both systems while determining the objective function and meeting the constraint on the product quality. The results provided by the dynamic model are necessary for the development of the optimal control structures. Thus in a future study, these dynamic models will be used to develop appropriate optimal control strategies for rejecting disturbances and to select proper start-up and shut-down procedures.

Control Strategy Designs and Simulations for a Biological Waste Water Treatment Process

A new and more appropriate continuous recycled system for aBiological Nutrient Removal process has been developed based on a sequencing batch reactor. This system comprises a Continuously Stirred Tank Reactor, a surge tank and a settling tank, from which a fraction of treated water is recycled back to the reactor. To design the control system for the whole plant, step tests have been conducted and Relative Gain Array analysis performed. Six control loops with the Process Variables including dissolved oxygen and nitrate concentrations, and volume holdups have been formed. Two designed control strategies Proportional Integral controllers and Generic Model Control have been implemented. The simulated results will be presented for comparison.