An optimisation-based algorithm for automated process flowsheet synthesis

Abstract

The automating of the process synthesis task that attains process flowsheet optimality is a new approach that can revolutionize the future of industrial plant design. This thesis is a contribution towards this approach; it presents an automated method for simultaneous structural and parametric optimisation of a process flowsheet. This algorithm is implemented for the design of a multi-component hollow-fibres membrane CO2 capture process flowsheet for a coal-fired power plant to achieve efficient carbon capture plant design. In this optimisation-based process flowsheet synthesis method, genetic algorithm plays the governing role in the design of a high-performance flowsheet. The role of GA is expanded by proposing and comparing a new crossover operator with four other commonly used operators. In this context, employing the random-point crossover, a 3.4% improvement in the value of objective function is observed over the simplest operator applied. Then, the potential for a hybrid process combining membrane and cryogenic separation is investigated to achieve an efficient design. The presented flowsheet could surpass membrane based CO2 capture system and in comparison, with other options, shows a promising possibility of exchanging conventional technologies with the proposed optimal configuration. Finally, a new optimisation algorithm is introduced termed as ‘Prenatal Genetic Screening’ Genetic Algorithm (PGS-GA) and its performance is compared against standard GA. This new evolutionary computation technique mimics the natural PGS procedure. PGS-GA leads to 2.3% improvement in the value of the objective function over the GA algorithm. The presence of repeated flowsheets among different solutions achieved using the algorithm starting from different randomly generated starting points that provide higher objective function values, approximately implies closeness of the solution to the global optimum.