The Optimal Design Of Seawater Desalination System Based On Genetic Algorithm

Seawater desalination technology is the main means of solving the problem of water shortage today, world water product is 40 million m3 daily, it has solved 100 million (1/40 of the world population) people's water supply. Seawater desalination technology has been grown up through development of nearly half century, and seawater desalination is getting more and more popular. But the bottleneck of further promoted and applied is the high cost of product water and lack of competence, so it is an important subject to decrease energy consumption of seawater desalination.Genetic algorithm is a random optimization method, whose operation object is the encoding of the parameters in the problem,namely chromosome, not the parameter itself, it can search in a large scale and easy to get the global optimum solution. Against the optimal design of seawater desalination system, this paper improves the conventional genetic algorithm, puts forward genetic algorithm with hybrid encoding which combines integer code with real code, the crossover and mutation operators are modified, it is solved by Visual C++6.0 program.The multi-stage flash(MSF) seawater desalination process is studied, the brine recycle MSF system is considered, three parameters are introduced which is the number of the heat recovery stage, the ratio between makeup water and the product water, and the ratio between the recycle brine and the feed seawater, based on the three parameters, the optimal design of MSF seawater desalination system is presented as a mixed-integer non-linear programming (MINLP), the novel optimum model of the process is established, the objective is the minimum water cost, and it is solved by improved genetic algorithm. An example is solved to the optimal design and the water cost obtained is less than that in the literature, the results are reasonable, it is proved that the models in this paper is effective and the solution approach put forward is feasible. The influence of the steam cost and energy cost on the water cost is also analyzed.Reverse osmosis (RO) seawater desalination process is studied; the detailed mathematical models of the units and the relevant economic models are introduced. The interdependence of the above models via some corresponding variables constitutes a system model. The membrane module can be classified for four types, i.e. higher salt-rejection membrane, high-flux membrane, normal membrane, lower salt-rejection membrane, taking the water cost as the objective, the design of the system can be presented as a mixed-integer non-linear programming (MINLP) considering thermodynamic, technical and flexibility constraints. Different optimal design options are obtained for different design specification and feed water quality.MSF-RO and RO-MSF hybrid seawater desalination systems are studied, MSF-RO takes the reject brine of the MSF process as the feed of the RO process while RO-MSF takes the reject brine of the RO process as the feed of the MSF process, the mixture of the RO and MSF product combines into the product water. Mathematical models are developed for the two hybrid systems, taking the minimum water cost as the objective, influence of different feed and output condition on the water cost is studied. Results show that given the output condition, when the feed concentration is lower, using the MSF-RO hybrid system is better, and else is RO-MSF; given the feed condition, when the output concentration is lower, using MSF-RO is better, and else is RO-MSF.