Composites of two perovskites are one possibility to combine protonic and p-type electronic conductivity as required for oxygen electrodes in protonic ceramic electrochemical cells. The BaCeO3-BaFeO3 system can be acceptor-doped to increase proton uptake and transport. However, preceding experiments [C. Berger et al., J. Mater. Chem. A 10 (2022) 2474; C. Nader et al., Solid State Ionics 406 (2024) 116474] indicated that the dopants are inhomogeneously distributed between the two phases, which is decisive for hydration ability and proton conductivity of such composites. Here, we use extended density functional theory calculations (DFT+U, Hubbard approach) for a comprehensive characterization of the BaCeO3-BaFeO3 system including acceptors. Supercells of various compositions are calculated to derive chemical reaction energies, for example for the transfer of defects between the phases. Two key aspects related to the hydration ability of such materials are: (i) The development of the electronic structure with increasing Fe content in a (hypothetical) single-phase BaCe1-xFexO3 perovskite. (ii) The distribution of acceptors (Ga3+, Sc3+, In3+, Y3+) and oxygen vacancies () between Ce- and Fe-rich phases. The segregation driving forces of acceptor dopant and are calculated individually. have the largest driving force towards the Fe-rich phase; ion radii and acid/base properties of the different acceptor dopants play a secondary role. The co-segregation of acceptors and into the ferrate phase unfortunately decreases the hydration ability of the Ce-rich proton conductor phase. Analogous trends are expected for related proton- and hole-conductor perovskite composites, which partially counteracts the intended mixed conductivity.