Search

In this paper, an integrated coal gasification combined cycle system for the production of hydrogen and electricity is optimized in terms of energy and exergy efficiencies, and the amount and cost of the produced hydrogen and electricity. The integrated system is optimized by focusing on the conversion process of coal to syngas. A novel optimization process is developed which integrates an artificial neural network with a genetic algorithm. The gasification system is modeled and simulated with Aspen Plus for large ranges of operating conditions, where the artificial neural network method is used to represent the simulation results mathematically. The mathematical model is then optimized using a genetic algorithm method. The optimization demonstrates that the lower is the grade of coal of the three considered coals, the less expensive is the hydrogen and electricity that can be produced by the considered integrated gasification combined cycle (IGCC) system. [ABSTRACT FROM AUTHOR]

Abstract: Applications for on-line data reconciliation and optimization must be efficient and numerically robust. The models in these applications are rarely changed and the same optimization problem is solved thousands of times with only minor changes in the parameters. This paper describes a suitable modeling framework for this type of applications that, with the aim of simplifying the creation of new models, makes the application robust and avoids numerical difficulties. The model is based on a unit model structure where first-order derivatives, scaling and initial values are properties of the unit model. A new scaling procedure is proposed based on equation and variable pairing. The modeling framework and the use of the proposed scaling procedure are demonstrated in two case studies, case 1 is simulation of a simple pipe model, case 2 is simulation, data reconciliation and optimization of a flash process. [Copyright &y& Elsevier]