This paper proposes a mixed-integer nonlinear programming model to define the best integration topology for a given set of renewable energy sources. The approach is based on the classical models of transmission network expansion planning, and defines a network topology for the connection of a set of small plants that minimizes the compromise between the investments in construction and the losses of energy over a given payback period. The methodology selects the best operational voltage levels for the lines, specify the substations and the circuit integration points, connecting the different energy sources. A demonstrative case study illustrates the benefits of the proposed methodology.