Model order reduction is targeting originally large-scale system models raise in many application problems as in physics and engineering. The main goal is producing an accurate, fast and cheap model that simulate the original large-scale one with high precision, with the ability to simulate as many different aspects of the full-order system under study, as possible. Reducing the calculation cost plays a main role in the offline phase, where all one-time calculations are done. As the partial differential equations govern optimal control systems; numerous computations of the optimality system are required; as an outcome of the discretization step, which can be then tackled by model order reduction. We are more interested in parametric model order reduction (PMOR), that is, dealing with problems for which the governing equations depend on a set of parameters with in a certain range.