Reduced superstructure solution of MINLP problem in refinery hydrogen management

Minimizing Hydrogen waste into fuel gas within the H2 network in a refinery is the objective function of an optimization problem in this paper. The superstructure obtained for a refinery wide concept, is first solved and validated for literature cases, then is reduced by heuristic rules, based on engineering judgment. The reduced superstructure contains all simulation procedures of pseudo-components definitions, fine tunings of all unit operations to reach actual operating conditions, reactions characterization, linear and nonlinear equalities and inequalities as system constraints. The set of governing equations are solved with Genetic Algorithm. Based on this optimization, in an Iranian refinery 22.6% reduction of H2 production and a saving of 1.19 million $/year could be achieved.