The reactive distillation and the reactive dividing wall column are examples of integrated reaction and separation processes, which show significant savings in investment as well as operational costs. Due to the high grade of integration the behavior of these processes is strongly nonlinear. This results in several local optima and thus the determination of energy optimal designs is quite challenging. Additionally, it is not trivial to understand and predict how the process performs in detail. Still, a fundamental process understanding is indispensable for a safe and energy efficient design of integrated reaction and separation processes. Therefore, a new hybrid global optimization method is presented, which is based on an evolutionary algorithm. This holistic approach is advantageous for optimization problems where the optimal values of the decision variables depend strongly on a certain decision variable. Dependencies of the process parameters can be deduced easily and a comprehensive process analysis can be carried out.