Optimal design of an electrodialysis brackish water desalination plant

This paper considers the optimal design and operation of electrodialysis (ED) desalination plants. In general an ED plant aims to produce potable water from a high salinity source, like brackish water or high salinity water. The system is modelled mathematically as mixed-integer non-linear programming (MINLP) optimization problem, determining the number of desalination stages, the membrane area, the total required energy so as to minimise the total annualised cost of the investment accounting for both infrastructure and operating costs. Two examples from the literature illustrate the applicability of the proposed approach and evaluate the quality of the results obtained.

This paper has introduced a detailed mathematical programming model that aims at addressing many complex issues of the design and operation problem of ED units. The overall problem was formulated as an MINLP model by minimising the total ED plant cost in terms of capital expenditure and running costs. A novel iterative optimisation-based solution procedure has been developed in contrast with the simulation approaches reported in the literature. Two illustrative examples have been tackled to demonstrate the applicability of the proposed approach. It can be proved that for the most economical operation the infrastructure and operational costs have to be equal [27]. In practice, the optimal conditions can almost never be met due to design and operational problems. The results from the two examples examined above show that the optimal solution is achieved when the infrastructure and operational costs are contributing almost equally. Since ED is a method that uses DC for the desalination, that makes the process attractive to be used in combination with renewable energy sources such as solar or wind power. Experimental runs have shown that photovoltaic systems can generate energy for systems that require up to 2 kWh/m3 to operate successfully [4]. Future work will explore the design option for ED plants based on renewable sources. Finally, it should be mentioned that mathematical programming is not a panacea, but it provides a reliable approximation of the required capabilities of the system and suggests ways to improve the performance of existing plants.