The goal of the project is to develop an accurate and computationally inexpensive method for calculating excitation energies of molecules. The new method will inherit low computational cost of density functionals and Hartree-Fock approach as well as proper description of valence excitations that DFT offers. A merge with a wavefunction theory will properly take into account other excitations unachievable with DFT itself, namely double excitations and charge-transfer excitations. The method will be proposed in the framework of Gross-Kohn-Oliveira ensemble minimum principle and would employ short-range density functionals. In particular, we would like to build a working program that calculates excitation energies with our method, validate it and test its performance. We would also like to study more closely the physical interpretation of Gross-Kohn-Oliveira ensemble minimum principle and the method itself.