Phase equilibrium calculations in systems subject to chemical reactions are involved in the design, synthesis and optimization of reactive separation processes. Until now, several methods have been developed to perform simultaneously physical and chemical equilibrium calculations. However, published methods may face numerical difficulties such as variable initialization dependence, divergence and convergence to trivial solutions or unstable equilibrium states. Besides, these methods generally use conventional composition variables and reactions extents as unknowns which directly affect the numerical implementation, reliability and efficiency of solving strategies. The objective of this work is to introduce and test an alternative approach to perform Gibbs energy minimization in phase equilibrium problems for reactive systems. Specifically, we have employed the transformed composition variables of Ung and Doherty and the stochastic optimization method Simulated Annealing for two-phase equilibrium calculations in reacting systems. Performance of this strategy has been tested using several benchmark problems and results show that proposed approach is generally suitable for the global minimization of transformed Gibbs energy in reactive systems with two-phase equilibrium.